Search Results for author: Ping Luo

Found 184 papers, 100 papers with code

Soft Neighbors are Positive Supporters in Contrastive Visual Representation Learning

no code implementations30 Mar 2023 Chongjian Ge, Jiangliu Wang, Zhan Tong, Shoufa Chen, Yibing Song, Ping Luo

We evaluate our soft neighbor contrastive learning method (SNCLR) on standard visual recognition benchmarks, including image classification, object detection, and instance segmentation.

DDP: Diffusion Model for Dense Visual Prediction

no code implementations30 Mar 2023 Yuanfeng Ji, Zhe Chen, Enze Xie, Lanqing Hong, Xihui Liu, Zhaoqiang Liu, Tong Lu, Zhenguo Li, Ping Luo

We propose a simple, efficient, yet powerful framework for dense visual predictions based on the conditional diffusion pipeline.

Real-time Controllable Denoising for Image and Video

no code implementations29 Mar 2023 Zhaoyang Zhang, Yitong Jiang, Wenqi Shao, Xiaogang Wang, Ping Luo, Kaimo Lin, Jinwei Gu

Controllable image denoising aims to generate clean samples with human perceptual priors and balance sharpness and smoothness.

Image Denoising Video Denoising

Accelerating Vision-Language Pretraining with Free Language Modeling

1 code implementation24 Mar 2023 Teng Wang, Yixiao Ge, Feng Zheng, Ran Cheng, Ying Shan, XiaoHu Qie, Ping Luo

FLM successfully frees the prediction rate from the tie-up with the corruption rate while allowing the corruption spans to be customized for each token to be predicted.

Language Modelling Masked Language Modeling

Dense Distinct Query for End-to-End Object Detection

1 code implementation22 Mar 2023 Shilong Zhang, Wang xinjiang, Jiaqi Wang, Jiangmiao Pang, Chengqi Lyu, Wenwei Zhang, Ping Luo, Kai Chen

Concretely, we first lay dense queries like traditional detectors and then select distinct ones for one-to-one assignments.

object-detection Object Detection

Vehicle-Infrastructure Cooperative 3D Object Detection via Feature Flow Prediction

1 code implementation19 Mar 2023 Haibao Yu, Yingjuan Tang, Enze Xie, Jilei Mao, Jirui Yuan, Ping Luo, Zaiqing Nie

Cooperatively utilizing both ego-vehicle and infrastructure sensor data can significantly enhance autonomous driving perception abilities.

3D Object Detection Autonomous Driving +1

Universal Instance Perception as Object Discovery and Retrieval

1 code implementation12 Mar 2023 Bin Yan, Yi Jiang, Jiannan Wu, Dong Wang, Ping Luo, Zehuan Yuan, Huchuan Lu

All instance perception tasks aim at finding certain objects specified by some queries such as category names, language expressions, and target annotations, but this complete field has been split into multiple independent subtasks.

 Ranked #1 on Referring Expression Comprehension on RefCOCOg-test (using extra training data)

Multi-Object Tracking and Segmentation Multiple Object Tracking +12

Learning Grounded Vision-Language Representation for Versatile Understanding in Untrimmed Videos

no code implementations11 Mar 2023 Teng Wang, Jinrui Zhang, Feng Zheng, Wenhao Jiang, Ran Cheng, Ping Luo

TEG learns to adaptively ground the possible event proposals given a set of sentences by estimating the cross-modal distance in a joint semantic space.

Dense Video Captioning Text Generation

AdaptDiffuser: Diffusion Models as Adaptive Self-evolving Planners

no code implementations3 Feb 2023 Zhixuan Liang, Yao Mu, Mingyu Ding, Fei Ni, Masayoshi Tomizuka, Ping Luo

For example, AdaptDiffuser not only outperforms the previous art Diffuser by 20. 8% on Maze2D and 7. 5% on MuJoCo locomotion, but also adapts better to new tasks, e. g., KUKA pick-and-place, by 27. 9% without requiring additional expert data.

Understanding Self-Supervised Pretraining with Part-Aware Representation Learning

no code implementations27 Jan 2023 Jie Zhu, Jiyang Qi, Mingyu Ding, Xiaokang Chen, Ping Luo, Xinggang Wang, Wenyu Liu, Leye Wang, Jingdong Wang

The study is mainly motivated by that random views, used in contrastive learning, and random masked (visible) patches, used in masked image modeling, are often about object parts.

Contrastive Learning Representation Learning

Fast-BEV: Towards Real-time On-vehicle Bird's-Eye View Perception

1 code implementation19 Jan 2023 Bin Huang, Yangguang Li, Enze Xie, Feng Liang, Luya Wang, Mingzhu Shen, Fenggang Liu, Tianqi Wang, Ping Luo, Jing Shao

Recently, the pure camera-based Bird's-Eye-View (BEV) perception removes expensive Lidar sensors, making it a feasible solution for economical autonomous driving.

Autonomous Driving Data Augmentation

Policy Adaptation from Foundation Model Feedback

no code implementations14 Dec 2022 Yuying Ge, Annabella Macaluso, Li Erran Li, Ping Luo, Xiaolong Wang

When deploying the trained policy to a new task or a new environment, we first let the policy play with randomly generated instructions to record the demonstrations.

Decision Making

MaskPlace: Fast Chip Placement via Reinforced Visual Representation Learning

no code implementations24 Nov 2022 Yao Lai, Yao Mu, Ping Luo

Firstly, MaskPlace recasts placement as a problem of learning pixel-level visual representation to comprehensively describe millions of modules on a chip, enabling placement in a high-resolution canvas and a large action space.

Layout Design Representation Learning

Prototypical context-aware dynamics generalization for high-dimensional model-based reinforcement learning

no code implementations23 Nov 2022 Junjie Wang, Yao Mu, Dong Li, Qichao Zhang, Dongbin Zhao, Yuzheng Zhuang, Ping Luo, Bin Wang, Jianye Hao

The latent world model provides a promising way to learn policies in a compact latent space for tasks with high-dimensional observations, however, its generalization across diverse environments with unseen dynamics remains challenging.

Model-based Reinforcement Learning reinforcement-learning +1

DiffusionDet: Diffusion Model for Object Detection

3 code implementations17 Nov 2022 Shoufa Chen, Peize Sun, Yibing Song, Ping Luo

In inference, the model refines a set of randomly generated boxes to the output results in a progressive way.

Denoising object-detection +1

Large-batch Optimization for Dense Visual Predictions

1 code implementation20 Oct 2022 Zeyue Xue, Jianming Liang, Guanglu Song, Zhuofan Zong, Liang Chen, Yu Liu, Ping Luo

To address this challenge, we propose a simple yet effective algorithm, named Adaptive Gradient Variance Modulator (AGVM), which can train dense visual predictors with very large batch size, enabling several benefits more appealing than prior arts.

Instance Segmentation object-detection +2

Decomposed Mutual Information Optimization for Generalized Context in Meta-Reinforcement Learning

1 code implementation9 Oct 2022 Yao Mu, Yuzheng Zhuang, Fei Ni, Bin Wang, Jianyu Chen, Jianye Hao, Ping Luo

This paper addresses such a challenge by Decomposed Mutual INformation Optimization (DOMINO) for context learning, which explicitly learns a disentangled context to maximize the mutual information between the context and historical trajectories, while minimizing the state transition prediction error.

Decision Making Meta Reinforcement Learning +2

Enhance Sample Efficiency and Robustness of End-to-end Urban Autonomous Driving via Semantic Masked World Model

no code implementations8 Oct 2022 Zeyu Gao, Yao Mu, Ruoyan Shen, Chen Chen, Yangang Ren, Jianyu Chen, Shengbo Eben Li, Ping Luo, YanFeng Lu

End-to-end autonomous driving provides a feasible way to automatically maximize overall driving system performance by directly mapping the raw pixels from a front-facing camera to control signals.

Autonomous Driving

Learning Transferable Spatiotemporal Representations from Natural Script Knowledge

1 code implementation30 Sep 2022 Ziyun Zeng, Yuying Ge, Xihui Liu, Bin Chen, Ping Luo, Shu-Tao Xia, Yixiao Ge

Pre-training on large-scale video data has become a common recipe for learning transferable spatiotemporal representations in recent years.

Representation Learning Video Understanding

FedVeca: Federated Vectorized Averaging on Non-IID Data with Adaptive Bi-directional Global Objective

no code implementations28 Sep 2022 Ping Luo, Jieren Cheng, Zhenhao Liu, N. Xiong, Jie Wu

However, the clients' Non-Independent and Identically Distributed (Non-IID) data negatively affect the trained model, and clients with different numbers of local updates may cause significant gaps to the local gradients in each communication round.

Federated Learning

Rethinking Resolution in the Context of Efficient Video Recognition

1 code implementation26 Sep 2022 Chuofan Ma, Qiushan Guo, Yi Jiang, Zehuan Yuan, Ping Luo, Xiaojuan Qi

Our key finding is that the major cause of degradation is not information loss in the down-sampling process, but rather the mismatch between network architecture and input scale.

Knowledge Distillation Video Recognition

ZoomNAS: Searching for Whole-body Human Pose Estimation in the Wild

1 code implementation23 Aug 2022 Lumin Xu, Sheng Jin, Wentao Liu, Chen Qian, Wanli Ouyang, Ping Luo, Xiaogang Wang

We propose a single-network approach, termed ZoomNet, to take into account the hierarchical structure of the full human body and solve the scale variation of different body parts.

2D Human Pose Estimation Neural Architecture Search +1

3D Interacting Hand Pose Estimation by Hand De-occlusion and Removal

1 code implementation22 Jul 2022 Hao Meng, Sheng Jin, Wentao Liu, Chen Qian, Mengxiang Lin, Wanli Ouyang, Ping Luo

Unlike most previous works that directly predict the 3D poses of two interacting hands simultaneously, we propose to decompose the challenging interacting hand pose estimation task and estimate the pose of each hand separately.

3D Interacting Hand Pose Estimation Hand Pose Estimation

Pose for Everything: Towards Category-Agnostic Pose Estimation

1 code implementation21 Jul 2022 Lumin Xu, Sheng Jin, Wang Zeng, Wentao Liu, Chen Qian, Wanli Ouyang, Ping Luo, Xiaogang Wang

In this paper, we introduce the task of Category-Agnostic Pose Estimation (CAPE), which aims to create a pose estimation model capable of detecting the pose of any class of object given only a few samples with keypoint definition.

Pose Estimation

Real-time End-to-End Video Text Spotter with Contrastive Representation Learning

no code implementations18 Jul 2022 Wejia Wu, Zhuang Li, Jiahong Li, Chunhua Shen, Hong Zhou, Size Li, Zhongyuan Wang, Ping Luo

Our contributions are three-fold: 1) CoText simultaneously address the three tasks (e. g., text detection, tracking, recognition) in a real-time end-to-end trainable framework.

Contrastive Learning Representation Learning +1

Towards Grand Unification of Object Tracking

1 code implementation14 Jul 2022 Bin Yan, Yi Jiang, Peize Sun, Dong Wang, Zehuan Yuan, Ping Luo, Huchuan Lu

We present a unified method, termed Unicorn, that can simultaneously solve four tracking problems (SOT, MOT, VOS, MOTS) with a single network using the same model parameters.

Multi-Object Tracking Multi-Object Tracking and Segmentation +2

Not All Models Are Equal: Predicting Model Transferability in a Self-challenging Fisher Space

1 code implementation7 Jul 2022 Wenqi Shao, Xun Zhao, Yixiao Ge, Zhaoyang Zhang, Lei Yang, Xiaogang Wang, Ying Shan, Ping Luo

It is challenging because the ground-truth model ranking for each task can only be generated by fine-tuning the pre-trained models on the target dataset, which is brute-force and computationally expensive.

Exploiting Context Information for Generic Event Boundary Captioning

1 code implementation3 Jul 2022 Jinrui Zhang, Teng Wang, Feng Zheng, Ran Cheng, Ping Luo

Previous methods only process the information of a single boundary at a time, which lacks utilization of video context information.

Boundary Captioning

VLMixer: Unpaired Vision-Language Pre-training via Cross-Modal CutMix

1 code implementation17 Jun 2022 Teng Wang, Wenhao Jiang, Zhichao Lu, Feng Zheng, Ran Cheng, Chengguo Yin, Ping Luo

Existing vision-language pre-training (VLP) methods primarily rely on paired image-text datasets, which are either annotated by enormous human labors, or crawled from the internet followed by elaborate data cleaning techniques.

Contrastive Learning Data Augmentation +1

CtrlFormer: Learning Transferable State Representation for Visual Control via Transformer

1 code implementation17 Jun 2022 Yao Mu, Shoufa Chen, Mingyu Ding, Jianyu Chen, Runjian Chen, Ping Luo

In visual control, learning transferable state representation that can transfer between different control tasks is important to reduce the training sample size.

Transfer Learning

AMOS: A Large-Scale Abdominal Multi-Organ Benchmark for Versatile Medical Image Segmentation

1 code implementation16 Jun 2022 Yuanfeng Ji, Haotian Bai, Jie Yang, Chongjian Ge, Ye Zhu, Ruimao Zhang, Zhen Li, Lingyan Zhang, Wanling Ma, Xiang Wan, Ping Luo

Constraint by the high cost of collecting and labeling 3D medical data, most of the deep learning models to date are driven by datasets with a limited number of organs of interest or samples, which still limits the power of modern deep models and makes it difficult to provide a fully comprehensive and fair estimate of various methods.

Image Segmentation Medical Image Segmentation +2

CO^3: Cooperative Unsupervised 3D Representation Learning for Autonomous Driving

1 code implementation8 Jun 2022 Runjian Chen, Yao Mu, Runsen Xu, Wenqi Shao, Chenhan Jiang, Hang Xu, Zhenguo Li, Ping Luo

In this paper, we propose CO^3, namely Cooperative Contrastive Learning and Contextual Shape Prediction, to learn 3D representation for outdoor-scene point clouds in an unsupervised manner.

Autonomous Driving Contrastive Learning +1

AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognition

1 code implementation26 May 2022 Shoufa Chen, Chongjian Ge, Zhan Tong, Jiangliu Wang, Yibing Song, Jue Wang, Ping Luo

To address this challenge, we propose an effective adaptation approach for Transformer, namely AdaptFormer, which can adapt the pre-trained ViTs into many different image and video tasks efficiently.

Action Recognition Video Recognition

Flow-based Recurrent Belief State Learning for POMDPs

no code implementations23 May 2022 Xiaoyu Chen, Yao Mu, Ping Luo, Shengbo Li, Jianyu Chen

Furthermore, we show that the learned belief states can be plugged into downstream RL algorithms to improve performance.

Decision Making Variational Inference

An Empirical Investigation of Representation Learning for Imitation

2 code implementations16 May 2022 Xin Chen, Sam Toyer, Cody Wild, Scott Emmons, Ian Fischer, Kuang-Huei Lee, Neel Alex, Steven H Wang, Ping Luo, Stuart Russell, Pieter Abbeel, Rohin Shah

We propose a modular framework for constructing representation learning algorithms, then use our framework to evaluate the utility of representation learning for imitation across several environment suites.

Image Classification Imitation Learning +1

MILES: Visual BERT Pre-training with Injected Language Semantics for Video-text Retrieval

1 code implementation26 Apr 2022 Yuying Ge, Yixiao Ge, Xihui Liu, Alex Jinpeng Wang, Jianping Wu, Ying Shan, XiaoHu Qie, Ping Luo

Dominant pre-training work for video-text retrieval mainly adopt the "dual-encoder" architectures to enable efficient retrieval, where two separate encoders are used to contrast global video and text representations, but ignore detailed local semantics.

Action Recognition Retrieval +5

Semantic-Aware Pretraining for Dense Video Captioning

no code implementations13 Apr 2022 Teng Wang, Zhu Liu, Feng Zheng, Zhichao Lu, Ran Cheng, Ping Luo

This report describes the details of our approach for the event dense-captioning task in ActivityNet Challenge 2021.

Dense Captioning Dense Video Captioning

M$^2$BEV: Multi-Camera Joint 3D Detection and Segmentation with Unified Birds-Eye View Representation

no code implementations11 Apr 2022 Enze Xie, Zhiding Yu, Daquan Zhou, Jonah Philion, Anima Anandkumar, Sanja Fidler, Ping Luo, Jose M. Alvarez

In this paper, we propose M$^2$BEV, a unified framework that jointly performs 3D object detection and map segmentation in the Birds Eye View~(BEV) space with multi-camera image inputs.

3D Object Detection object-detection

DaViT: Dual Attention Vision Transformers

3 code implementations7 Apr 2022 Mingyu Ding, Bin Xiao, Noel Codella, Ping Luo, Jingdong Wang, Lu Yuan

We show that these two self-attentions complement each other: (i) since each channel token contains an abstract representation of the entire image, the channel attention naturally captures global interactions and representations by taking all spatial positions into account when computing attention scores between channels; (ii) the spatial attention refines the local representations by performing fine-grained interactions across spatial locations, which in turn helps the global information modeling in channel attention.

Ranked #11 on Image Classification on ImageNet (using extra training data)

Image Classification Instance Segmentation +2

Scale-Equivalent Distillation for Semi-Supervised Object Detection

no code implementations CVPR 2022 Qiushan Guo, Yao Mu, Jianyu Chen, Tianqi Wang, Yizhou Yu, Ping Luo

Further, we overcome these challenges by introducing a novel approach, Scale-Equivalent Distillation (SED), which is a simple yet effective end-to-end knowledge distillation framework robust to large object size variance and class imbalance.

Knowledge Distillation object-detection +2

Compression of Generative Pre-trained Language Models via Quantization

no code implementations ACL 2022 Chaofan Tao, Lu Hou, Wei zhang, Lifeng Shang, Xin Jiang, Qun Liu, Ping Luo, Ngai Wong

We find that previous quantization methods fail on generative tasks due to the \textit{homogeneous word embeddings} caused by reduced capacity, and \textit{varied distribution of weights}.

Model Compression Quantization +1

End-to-End Video Text Spotting with Transformer

1 code implementation20 Mar 2022 Weijia Wu, Yuanqiang Cai, Chunhua Shen, Debing Zhang, Ying Fu, Hong Zhou, Ping Luo

Recent video text spotting methods usually require the three-staged pipeline, i. e., detecting text in individual images, recognizing localized text, tracking text streams with post-processing to generate final results.

Text Spotting

WegFormer: Transformers for Weakly Supervised Semantic Segmentation

no code implementations16 Mar 2022 Chunmeng Liu, Enze Xie, Wenjia Wang, Wenhai Wang, Guangyao Li, Ping Luo

Although convolutional neural networks (CNNs) have achieved remarkable progress in weakly supervised semantic segmentation (WSSS), the effective receptive field of CNN is insufficient to capture global context information, leading to sub-optimal results.

Weakly supervised Semantic Segmentation Weakly-Supervised Semantic Segmentation

Context Autoencoder for Self-Supervised Representation Learning

4 code implementations7 Feb 2022 Xiaokang Chen, Mingyu Ding, Xiaodi Wang, Ying Xin, Shentong Mo, Yunhao Wang, Shumin Han, Ping Luo, Gang Zeng, Jingdong Wang

We introduce an alignment constraint, encouraging that the representations for masked patches, predicted from the encoded representations of visible patches, are aligned with the masked patch presentations computed from the encoder.

Instance Segmentation object-detection +5

Pseudo-Labeled Auto-Curriculum Learning for Semi-Supervised Keypoint Localization

no code implementations ICLR 2022 Can Wang, Sheng Jin, Yingda Guan, Wentao Liu, Chen Qian, Ping Luo, Wanli Ouyang

PL approaches apply pseudo-labels to unlabeled data, and then train the model with a combination of the labeled and pseudo-labeled data iteratively.

MetaDance: Few-shot Dancing Video Retargeting via Temporal-aware Meta-learning

no code implementations13 Jan 2022 Yuying Ge, Yibing Song, Ruimao Zhang, Ping Luo

Dancing video retargeting aims to synthesize a video that transfers the dance movements from a source video to a target person.


Bridging Video-text Retrieval with Multiple Choice Questions

2 code implementations CVPR 2022 Yuying Ge, Yixiao Ge, Xihui Liu, Dian Li, Ying Shan, XiaoHu Qie, Ping Luo

As an additional benefit, our method achieves competitive results with much shorter pre-training videos on single-modality downstream tasks, e. g., action recognition with linear evaluation.

Ranked #32 on Video Retrieval on MSR-VTT-1kA (using extra training data)

Action Recognition Multiple-choice +7

Language as Queries for Referring Video Object Segmentation

1 code implementation CVPR 2022 Jiannan Wu, Yi Jiang, Peize Sun, Zehuan Yuan, Ping Luo

Referring video object segmentation (R-VOS) is an emerging cross-modal task that aims to segment the target object referred by a language expression in all video frames.

 Ranked #1 on Referring Expression Segmentation on A2D Sentences (using extra training data)

Object Tracking Referring Expression Segmentation +4

MetaCloth: Learning Unseen Tasks of Dense Fashion Landmark Detection from a Few Samples

no code implementations6 Dec 2021 Yuying Ge, Ruimao Zhang, Ping Luo

This work proposes a novel framework named MetaCloth via meta-learning, which is able to learn unseen tasks of dense fashion landmark detection with only a few annotated samples.


Compressed Video Contrastive Learning

no code implementations NeurIPS 2021 Yuqi Huo, Mingyu Ding, Haoyu Lu, Nanyi Fei, Zhiwu Lu, Ji-Rong Wen, Ping Luo

To enhance the representation ability of the motion vectors, hence the effectiveness of our method, we design a cross guidance contrastive learning algorithm based on multi-instance InfoNCE loss, where motion vectors can take supervision signals from RGB frames and vice versa.

Contrastive Learning Representation Learning

Revitalizing CNN Attention via Transformers in Self-Supervised Visual Representation Learning

1 code implementation NeurIPS 2021 Chongjian Ge, Youwei Liang, Yibing Song, Jianbo Jiao, Jue Wang, Ping Luo

Motivated by the transformers that explore visual attention effectively in recognition scenarios, we propose a CNN Attention REvitalization (CARE) framework to train attentive CNN encoders guided by transformers in SSL.

Image Classification object-detection +3

DanceTrack: Multi-Object Tracking in Uniform Appearance and Diverse Motion

3 code implementations CVPR 2022 Peize Sun, Jinkun Cao, Yi Jiang, Zehuan Yuan, Song Bai, Kris Kitani, Ping Luo

A typical pipeline for multi-object tracking (MOT) is to use a detector for object localization, and following re-identification (re-ID) for object association.

Association Multi-Object Tracking +3

FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation

2 code implementations3 Nov 2021 Zhe Chen, Jiahao Wang, Wenhai Wang, Guo Chen, Enze Xie, Ping Luo, Tong Lu

We propose an accurate and efficient scene text detection framework, termed FAST (i. e., faster arbitrarily-shaped text detector).

Image Classification Scene Text Detection

Dynamic Visual Reasoning by Learning Differentiable Physics Models from Video and Language

no code implementations NeurIPS 2021 Mingyu Ding, Zhenfang Chen, Tao Du, Ping Luo, Joshua B. Tenenbaum, Chuang Gan

This is achieved by seamlessly integrating three components: a visual perception module, a concept learner, and a differentiable physics engine.

Visual Reasoning

Revitalizing CNN Attentions via Transformers in Self-Supervised Visual Representation Learning

1 code implementation11 Oct 2021 Chongjian Ge, Youwei Liang, Yibing Song, Jianbo Jiao, Jue Wang, Ping Luo

Motivated by the transformers that explore visual attention effectively in recognition scenarios, we propose a CNN Attention REvitalization (CARE) framework to train attentive CNN encoders guided by transformers in SSL.

Image Classification object-detection +3

Objects in Semantic Topology

no code implementations ICLR 2022 Shuo Yang, Peize Sun, Yi Jiang, Xiaobo Xia, Ruiheng Zhang, Zehuan Yuan, Changhu Wang, Ping Luo, Min Xu

A more realistic object detection paradigm, Open-World Object Detection, has arisen increasing research interests in the community recently.

Incremental Learning Language Modelling +2

Scale-Invariant Teaching for Semi-Supervised Object Detection

no code implementations29 Sep 2021 Qiushan Guo, Yizhou Yu, Ping Luo

Furthermore, the limited annotations in semi-supervised learning scale up the challenges: large variance of object sizes and class imbalance (i. e., the extreme ratio between background and object), hindering the performance of prior arts.

object-detection Pseudo Label

Towards High-Quality Temporal Action Detection with Sparse Proposals

1 code implementation18 Sep 2021 Jiannan Wu, Peize Sun, Shoufa Chen, Jiewen Yang, Zihao Qi, Lan Ma, Ping Luo

Towards high-quality temporal action detection, we introduce Sparse Proposals to interact with the hierarchical features.

Action Detection Avg +1

Adversarial Robustness for Unsupervised Domain Adaptation

no code implementations ICCV 2021 Muhammad Awais, Fengwei Zhou, Hang Xu, Lanqing Hong, Ping Luo, Sung-Ho Bae, Zhenguo Li

Extensive Unsupervised Domain Adaptation (UDA) studies have shown great success in practice by learning transferable representations across a labeled source domain and an unlabeled target domain with deep models.

Adversarial Robustness Unsupervised Domain Adaptation

CycleMLP: A MLP-like Architecture for Dense Prediction

8 code implementations ICLR 2022 Shoufa Chen, Enze Xie, Chongjian Ge, Runjian Chen, Ding Liang, Ping Luo

We build a family of models which surpass existing MLPs and even state-of-the-art Transformer-based models, e. g., Swin Transformer, while using fewer parameters and FLOPs.

Image Classification Instance Segmentation +3

Multi-frame Collaboration for Effective Endoscopic Video Polyp Detection via Spatial-Temporal Feature Transformation

1 code implementation8 Jul 2021 Lingyun Wu, Zhiqiang Hu, Yuanfeng Ji, Ping Luo, Shaoting Zhang

For example, STFT improves the still image baseline FCOS by 10. 6% and 20. 6% on the comprehensive F1-score of the polyp localization task in CVC-Clinic and ASUMayo datasets, respectively, and outperforms the state-of-the-art video-based method by 3. 6% and 8. 0%, respectively.

Multi-Compound Transformer for Accurate Biomedical Image Segmentation

1 code implementation28 Jun 2021 Yuanfeng Ji, Ruimao Zhang, Huijie Wang, Zhen Li, Lingyun Wu, Shaoting Zhang, Ping Luo

The recent vision transformer(i. e. for image classification) learns non-local attentive interaction of different patch tokens.

Image Classification Image Segmentation +2

HR-NAS: Searching Efficient High-Resolution Neural Architectures with Lightweight Transformers

1 code implementation CVPR 2021 Mingyu Ding, Xiaochen Lian, Linjie Yang, Peng Wang, Xiaojie Jin, Zhiwu Lu, Ping Luo

Last, we proposed an efficient fine-grained search strategy to train HR-NAS, which effectively explores the search space, and finds optimal architectures given various tasks and computation resources.

Image Classification Neural Architecture Search +2

SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers

14 code implementations NeurIPS 2021 Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo

We present SegFormer, a simple, efficient yet powerful semantic segmentation framework which unifies Transformers with lightweight multilayer perception (MLP) decoders.

Semantic Segmentation Thermal Image Segmentation

Extracting Variable-Depth Logical Document Hierarchy from Long Documents: Method, Evaluation, and Application

no code implementations14 May 2021 Rongyu Cao, Yixuan Cao, Ganbin Zhou, Ping Luo

In this paper, we study the problem of extracting variable-depth "logical document hierarchy" from long documents, namely organizing the recognized "physical document objects" into hierarchical structures.

Passage Retrieval Retrieval

When Human Pose Estimation Meets Robustness: Adversarial Algorithms and Benchmarks

1 code implementation CVPR 2021 Jiahang Wang, Sheng Jin, Wentao Liu, Weizhong Liu, Chen Qian, Ping Luo

However, unlike human vision that is robust to various data corruptions such as blur and pixelation, current pose estimators are easily confused by these corruptions.

Knowledge Distillation Pose Estimation

BWCP: Probabilistic Learning-to-Prune Channels for ConvNets via Batch Whitening

no code implementations13 May 2021 Wenqi Shao, Hang Yu, Zhaoyang Zhang, Hang Xu, Zhenguo Li, Ping Luo

To address this problem, we develop a probability-based pruning algorithm, called batch whitening channel pruning (BWCP), which can stochastically discard unimportant channels by modeling the probability of a channel being activated.

PolarMask++: Enhanced Polar Representation for Single-Shot Instance Segmentation and Beyond

1 code implementation5 May 2021 Enze Xie, Wenhai Wang, Mingyu Ding, Ruimao Zhang, Ping Luo

Extensive experiments demonstrate the effectiveness of both PolarMask and PolarMask++, which achieve competitive results on instance segmentation in the challenging COCO dataset with single-model and single-scale training and testing, as well as new state-of-the-art results on rotate text detection and cell segmentation.

Ranked #77 on Instance Segmentation on COCO test-dev (using extra training data)

Cell Segmentation Instance Segmentation +3

Going Deeper Into Face Detection: A Survey

no code implementations27 Mar 2021 Shervin Minaee, Ping Luo, Zhe Lin, Kevin Bowyer

In this work, we provide a detailed overview of some of the most representative deep learning based face detection methods by grouping them into a few major categories, and present their core architectural designs and accuracies on popular benchmarks.

Face Detection Image Classification

Learning Versatile Neural Architectures by Propagating Network Codes

1 code implementation ICLR 2022 Mingyu Ding, Yuqi Huo, Haoyu Lu, Linjie Yang, Zhe Wang, Zhiwu Lu, Jingdong Wang, Ping Luo

(4) Thorough studies of NCP on inter-, cross-, and intra-tasks highlight the importance of cross-task neural architecture design, i. e., multitask neural architectures and architecture transferring between different tasks.

Image Segmentation Neural Architecture Search +2

Towards Ultra-Resolution Neural Style Transfer via Thumbnail Instance Normalization

1 code implementation22 Mar 2021 Zhe Chen, Wenhai Wang, Enze Xie, Tong Lu, Ping Luo

(1) We divide input image into small patches and adopt TIN, successfully transferring image style with arbitrary high-resolution.

Style Transfer

Disentangled Cycle Consistency for Highly-realistic Virtual Try-On

1 code implementation CVPR 2021 Chongjian Ge, Yibing Song, Yuying Ge, Han Yang, Wei Liu, Ping Luo

To this end, DCTON can be naturally trained in a self-supervised manner following cycle consistency learning.

Virtual Try-on

Parser-Free Virtual Try-on via Distilling Appearance Flows

1 code implementation CVPR 2021 Yuying Ge, Yibing Song, Ruimao Zhang, Chongjian Ge, Wei Liu, Ping Luo

A recent pioneering work employed knowledge distillation to reduce the dependency of human parsing, where the try-on images produced by a parser-based method are used as supervisions to train a "student" network without relying on segmentation, making the student mimic the try-on ability of the parser-based model.

Human Parsing Knowledge Distillation +1

Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions

9 code implementations ICCV 2021 Wenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan, Kaitao Song, Ding Liang, Tong Lu, Ping Luo, Ling Shao

Unlike the recently-proposed Transformer model (e. g., ViT) that is specially designed for image classification, we propose Pyramid Vision Transformer~(PVT), which overcomes the difficulties of porting Transformer to various dense prediction tasks.

Image Classification Instance Segmentation +3

FAT: Learning Low-Bitwidth Parametric Representation via Frequency-Aware Transformation

1 code implementation15 Feb 2021 Chaofan Tao, Rui Lin, Quan Chen, Zhaoyang Zhang, Ping Luo, Ngai Wong

Prior arts often discretize the network weights by carefully tuning hyper-parameters of quantization (e. g. non-uniform stepsize and layer-wise bitwidths), which are complicated and sub-optimal because the full-precision and low-precision models have a large discrepancy.

Neural Network Compression Quantization

DetCo: Unsupervised Contrastive Learning for Object Detection

2 code implementations ICCV 2021 Enze Xie, Jian Ding, Wenhai Wang, Xiaohang Zhan, Hang Xu, Peize Sun, Zhenguo Li, Ping Luo

Unlike most recent methods that focused on improving accuracy of image classification, we present a novel contrastive learning approach, named DetCo, which fully explores the contrasts between global image and local image patches to learn discriminative representations for object detection.

Contrastive Learning Image Classification +1

Segmenting Transparent Object in the Wild with Transformer

2 code implementations21 Jan 2021 Enze Xie, Wenjia Wang, Wenhai Wang, Peize Sun, Hang Xu, Ding Liang, Ping Luo

This work presents a new fine-grained transparent object segmentation dataset, termed Trans10K-v2, extending Trans10K-v1, the first large-scale transparent object segmentation dataset.

Semantic Segmentation Transparent objects

Rethinking the Pruning Criteria for Convolutional Neural Network

no code implementations NeurIPS 2021 Zhongzhan Huang, Xinjiang Wang, Ping Luo

Channel pruning is a popular technique for compressing convolutional neural networks (CNNs), and various pruning criteria have been proposed to remove the redundant filters of CNNs.

Self-Supervised Video Representation Learning with Constrained Spatiotemporal Jigsaw

no code implementations1 Jan 2021 Yuqi Huo, Mingyu Ding, Haoyu Lu, Zhiwu Lu, Tao Xiang, Ji-Rong Wen, Ziyuan Huang, Jianwen Jiang, Shiwei Zhang, Mingqian Tang, Songfang Huang, Ping Luo

With the constrained jigsaw puzzles, instead of solving them directly, which could still be extremely hard, we carefully design four surrogate tasks that are more solvable but meanwhile still ensure that the learned representation is sensitive to spatiotemporal continuity at both the local and global levels.

Representation Learning

Bringing Events Into Video Deblurring With Non-Consecutively Blurry Frames

1 code implementation ICCV 2021 Wei Shang, Dongwei Ren, Dongqing Zou, Jimmy S. Ren, Ping Luo, WangMeng Zuo

EFM can also be easily incorporated into existing deblurring networks, making event-driven deblurring task benefit from state-of-the-art deblurring methods.


What Makes for End-to-End Object Detection?

1 code implementation10 Dec 2020 Peize Sun, Yi Jiang, Enze Xie, Wenqi Shao, Zehuan Yuan, Changhu Wang, Ping Luo

We identify that classification cost in matching cost is the main ingredient: (1) previous detectors only consider location cost, (2) by additionally introducing classification cost, previous detectors immediately produce one-to-one prediction during inference.

General Classification object-detection +1

Polygon-free: Unconstrained Scene Text Detection with Box Annotations

1 code implementation26 Nov 2020 Weijia Wu, Enze Xie, Ruimao Zhang, Wenhai Wang, Hong Zhou, Ping Luo

For example, without using polygon annotations, PSENet achieves an 80. 5% F-score on TotalText [3] (vs. 80. 9% of fully supervised counterpart), 31. 1% better than training directly with upright bounding box annotations, and saves 80%+ labeling costs.

Scene Text Detection

Sparse R-CNN: End-to-End Object Detection with Learnable Proposals

6 code implementations CVPR 2021 Peize Sun, Rufeng Zhang, Yi Jiang, Tao Kong, Chenfeng Xu, Wei Zhan, Masayoshi Tomizuka, Lei LI, Zehuan Yuan, Changhu Wang, Ping Luo

In our method, however, a fixed sparse set of learned object proposals, total length of $N$, are provided to object recognition head to perform classification and location.

object-detection Object Detection +1

Do 2D GANs Know 3D Shape? Unsupervised 3D shape reconstruction from 2D Image GANs

1 code implementation ICLR 2021 Xingang Pan, Bo Dai, Ziwei Liu, Chen Change Loy, Ping Luo

Through our investigation, we found that such a pre-trained GAN indeed contains rich 3D knowledge and thus can be used to recover 3D shape from a single 2D image in an unsupervised manner.

3D Shape Reconstruction

UXNet: Searching Multi-level Feature Aggregation for 3D Medical Image Segmentation

no code implementations16 Sep 2020 Yuanfeng Ji, Ruimao Zhang, Zhen Li, Jiamin Ren, Shaoting Zhang, Ping Luo

Unlike the recent neural architecture search (NAS) methods that typically searched the optimal operators in each network layer, but missed a good strategy to search for feature aggregations, this paper proposes a novel NAS method for 3D medical image segmentation, named UXNet, which searches both the scale-wise feature aggregation strategies as well as the block-wise operators in the encoder-decoder network.

Image Segmentation Neural Architecture Search +2

RelativeNAS: Relative Neural Architecture Search via Slow-Fast Learning

2 code implementations14 Sep 2020 Hao Tan, Ran Cheng, Shihua Huang, Cheng He, Changxiao Qiu, Fan Yang, Ping Luo

Despite the remarkable successes of Convolutional Neural Networks (CNNs) in computer vision, it is time-consuming and error-prone to manually design a CNN.

Keypoint Detection Neural Architecture Search +3

Compensation Tracker: Reprocessing Lost Object for Multi-Object Tracking

no code implementations27 Aug 2020 Zhibo Zou, Jun-Jie Huang, Ping Luo

Based on simple and traditional methods, we propose a compensation tracker to further alleviate the lost tracking problem caused by missing detection.

Motion Compensation Multi-Object Tracking

Dynamic and Static Context-aware LSTM for Multi-agent Motion Prediction

no code implementations ECCV 2020 Chaofan Tao, Qinhong Jiang, Lixin Duan, Ping Luo

Existing work addressed this challenge by either learning social spatial interactions represented by the positions of a group of pedestrians, while ignoring their temporal coherence (\textit{i. e.} dependencies between different long trajectories), or by understanding the complicated scene layout (\textit{e. g.} scene segmentation) to ensure safe navigation.

motion prediction Trajectory Prediction

AE TextSpotter: Learning Visual and Linguistic Representation for Ambiguous Text Spotting

2 code implementations ECCV 2020 Wenhai Wang, Xuebo Liu, Xiaozhong Ji, Enze Xie, Ding Liang, Zhibo Yang, Tong Lu, Chunhua Shen, Ping Luo

Unlike previous works that merely employed visual features for text detection, this work proposes a novel text spotter, named Ambiguity Eliminating Text Spotter (AE TextSpotter), which learns both visual and linguistic features to significantly reduce ambiguity in text detection.

Language Modelling Text Spotting

Whole-Body Human Pose Estimation in the Wild

2 code implementations ECCV 2020 Sheng Jin, Lumin Xu, Jin Xu, Can Wang, Wentao Liu, Chen Qian, Wanli Ouyang, Ping Luo

This paper investigates the task of 2D human whole-body pose estimation, which aims to localize dense landmarks on the entire human body including face, hands, body, and feet.

2D Human Pose Estimation Facial Landmark Detection +2

Differentiable Hierarchical Graph Grouping for Multi-Person Pose Estimation

no code implementations ECCV 2020 Sheng Jin, Wentao Liu, Enze Xie, Wenhai Wang, Chen Qian, Wanli Ouyang, Ping Luo

The modules of HGG can be trained end-to-end with the keypoint detection network and is able to supervise the grouping process in a hierarchical manner.

2D Human Pose Estimation Graph Clustering +3

3D Human Mesh Regression with Dense Correspondence

3 code implementations CVPR 2020 Wang Zeng, Wanli Ouyang, Ping Luo, Wentao Liu, Xiaogang Wang

This paper proposes a model-free 3D human mesh estimation framework, named DecoMR, which explicitly establishes the dense correspondence between the mesh and the local image features in the UV space (i. e. a 2D space used for texture mapping of 3D mesh).

3D Human Pose Estimation 3D Human Reconstruction +1

Learning a Reinforced Agent for Flexible Exposure Bracketing Selection

1 code implementation CVPR 2020 Zhouxia Wang, Jiawei Zhang, Mude Lin, Jiong Wang, Ping Luo, Jimmy Ren

Automatically selecting exposure bracketing (images exposed differently) is important to obtain a high dynamic range image by using multi-exposure fusion.

Convolution-Weight-Distribution Assumption: Rethinking the Criteria of Channel Pruning

no code implementations24 Apr 2020 Zhongzhan Huang, Wenqi Shao, Xinjiang Wang, Liang Lin, Ping Luo

Channel pruning is a popular technique for compressing convolutional neural networks (CNNs), where various pruning criteria have been proposed to remove the redundant filters.

AdaX: Adaptive Gradient Descent with Exponential Long Term Memory

1 code implementation21 Apr 2020 Wenjie Li, Zhaoyang Zhang, Xinjiang Wang, Ping Luo

Although adaptive optimization algorithms such as Adam show fast convergence in many machine learning tasks, this paper identifies a problem of Adam by analyzing its performance in a simple non-convex synthetic problem, showing that Adam's fast convergence would possibly lead the algorithm to local minimums.

Segmenting Transparent Objects in the Wild

1 code implementation ECCV 2020 Enze Xie, Wenjia Wang, Wenhai Wang, Mingyu Ding, Chunhua Shen, Ping Luo

To address this important problem, this work proposes a large-scale dataset for transparent object segmentation, named Trans10K, consisting of 10, 428 images of real scenarios with carefully manual annotations, which are 10 times larger than the existing datasets.

Semantic Segmentation Transparent objects

Domain-Adaptive Few-Shot Learning

1 code implementation19 Mar 2020 An Zhao, Mingyu Ding, Zhiwu Lu, Tao Xiang, Yulei Niu, Jiechao Guan, Ji-Rong Wen, Ping Luo

Existing few-shot learning (FSL) methods make the implicit assumption that the few target class samples are from the same domain as the source class samples.

Domain Adaptation Few-Shot Learning

Exemplar Normalization for Learning Deep Representation

no code implementations CVPR 2020 Ruimao Zhang, Zhanglin Peng, Lingyun Wu, Zhen Li, Ping Luo

This work investigates a novel dynamic learning-to-normalize (L2N) problem by proposing Exemplar Normalization (EN), which is able to learn different normalization methods for different convolutional layers and image samples of a deep network.

Semantic Segmentation

Towards Photo-Realistic Virtual Try-On by Adaptively Generating$\leftrightarrow$Preserving Image Content

3 code implementations12 Mar 2020 Han Yang, Ruimao Zhang, Xiaobao Guo, Wei Liu, WangMeng Zuo, Ping Luo

First, a semantic layout generation module utilizes semantic segmentation of the reference image to progressively predict the desired semantic layout after try-on.

Ranked #4 on Virtual Try-on on VITON (IS metric)

Semantic Segmentation Virtual Try-on

Channel Equilibrium Networks for Learning Deep Representation

1 code implementation ICML 2020 Wenqi Shao, Shitao Tang, Xingang Pan, Ping Tan, Xiaogang Wang, Ping Luo

Unlike prior arts that simply removed the inhibited channels, we propose to "wake them up" during training by designing a novel neural building block, termed Channel Equilibrium (CE) block, which enables channels at the same layer to contribute equally to the learned representation.

How Does BN Increase Collapsed Neural Network Filters?

no code implementations30 Jan 2020 Sheng Zhou, Xinjiang Wang, Ping Luo, Litong Feng, Wenjie Li, Wei zhang

This phenomenon is caused by the normalization effect of BN, which induces a non-trainable region in the parameter space and reduces the network capacity as a result.

object-detection Object Detection

Every Frame Counts: Joint Learning of Video Segmentation and Optical Flow

no code implementations28 Nov 2019 Mingyu Ding, Zhe Wang, Bolei Zhou, Jianping Shi, Zhiwu Lu, Ping Luo

Moreover, our framework is able to utilize both labeled and unlabeled frames in the video through joint training, while no additional calculation is required in inference.

Optical Flow Estimation Semantic Segmentation +2

Vision-Infused Deep Audio Inpainting

no code implementations ICCV 2019 Hang Zhou, Ziwei Liu, Xudong Xu, Ping Luo, Xiaogang Wang

Extensive experiments demonstrate that our framework is capable of inpainting realistic and varying audio segments with or without visual contexts.

Audio inpainting Image Inpainting

PolarMask: Single Shot Instance Segmentation with Polar Representation

2 code implementations CVPR 2020 Enze Xie, Peize Sun, Xiaoge Song, Wenhai Wang, Ding Liang, Chunhua Shen, Ping Luo

In this paper, we introduce an anchor-box free and single shot instance segmentation method, which is conceptually simple, fully convolutional and can be used as a mask prediction module for instance segmentation, by easily embedding it into most off-the-shelf detection methods.

Instance Segmentation Object Detection +2

Channel Equilibrium Networks

no code implementations25 Sep 2019 Wenqi Shao, Shitao Tang, Xingang Pan, Ping Tan, Xiaogang Wang, Ping Luo

However, over-sparse CNNs have many collapsed channels (i. e. many channels with undesired zero values), impeding their learning ability.

TextSR: Content-Aware Text Super-Resolution Guided by Recognition

1 code implementation16 Sep 2019 Wenjia Wang, Enze Xie, Peize Sun, Wenhai Wang, Lixun Tian, Chunhua Shen, Ping Luo

Nonetheless, most of the previous methods may not work well in recognizing text with low resolution which is often seen in natural scene images.

Scene Text Recognition Super-Resolution

PDA: Progressive Data Augmentation for General Robustness of Deep Neural Networks

no code implementations11 Sep 2019 Hang Yu, Aishan Liu, Xianglong Liu, Gengchao Li, Ping Luo, Ran Cheng, Jichen Yang, Chongzhi Zhang

In other words, DNNs trained with PDA are able to obtain more robustness against both adversarial attacks as well as common corruptions than the recent state-of-the-art methods.

Data Augmentation

Scale Calibrated Training: Improving Generalization of Deep Networks via Scale-Specific Normalization

no code implementations31 Aug 2019 Zhuoran Yu, Aojun Zhou, Yukun Ma, Yudian Li, Xiaohan Zhang, Ping Luo

Experiment results show that SCT improves accuracy of single Resnet-50 on ImageNet by 1. 7% and 11. 5% accuracy when testing on image sizes of 224 and 128 respectively.

Data Augmentation Image Classification +1

Fashion Retrieval via Graph Reasoning Networks on a Similarity Pyramid

no code implementations ICCV 2019 Zhanghui Kuang, Yiming Gao, Guanbin Li, Ping Luo, Yimin Chen, Liang Lin, Wayne Zhang

To address this issue, we propose a novel Graph Reasoning Network (GRNet) on a Similarity Pyramid, which learns similarities between a query and a gallery cloth by using both global and local representations in multiple scales.

Image Retrieval Retrieval

Differentiable Learning-to-Group Channels via Groupable Convolutional Neural Networks

no code implementations ICCV 2019 Zhaoyang Zhang, Jingyu Li, Wenqi Shao, Zhanglin Peng, Ruimao Zhang, Xiaogang Wang, Ping Luo

ResNeXt, still suffers from the sub-optimal performance due to manually defining the number of groups as a constant over all of the layers.

Once a MAN: Towards Multi-Target Attack via Learning Multi-Target Adversarial Network Once

no code implementations ICCV 2019 Jiangfan Han, Xiaoyi Dong, Ruimao Zhang, Dong-Dong Chen, Weiming Zhang, Nenghai Yu, Ping Luo, Xiaogang Wang

Recently, generation-based methods have received much attention since they directly use feed-forward networks to generate the adversarial samples, which avoid the time-consuming iterative attacking procedure in optimization-based and gradient-based methods.

Classification General Classification

Deep Self-Learning From Noisy Labels

no code implementations ICCV 2019 Jiangfan Han, Ping Luo, Xiaogang Wang

Unlike previous works constrained by many conditions, making them infeasible to real noisy cases, this work presents a novel deep self-learning framework to train a robust network on the real noisy datasets without extra supervision.

Learning with noisy labels Self-Learning

MaskGAN: Towards Diverse and Interactive Facial Image Manipulation

6 code implementations CVPR 2020 Cheng-Han Lee, Ziwei Liu, Lingyun Wu, Ping Luo

To overcome these drawbacks, we propose a novel framework termed MaskGAN, enabling diverse and interactive face manipulation.

Image Manipulation

Switchable Normalization for Learning-to-Normalize Deep Representation

no code implementations22 Jul 2019 Ping Luo, Ruimao Zhang, Jiamin Ren, Zhanglin Peng, Jingyu Li

Analyses of SN are also presented to answer the following three questions: (a) Is it useful to allow each normalization layer to select its own normalizer?

Atom Responding Machine for Dialog Generation

no code implementations14 May 2019 Ganbin Zhou, Ping Luo, Jingwu Chen, Fen Lin, Leyu Lin, Qing He

To enrich the generated responses, ARM introduces a large number of molecule-mechanisms as various responding styles, which are conducted by taking different combinations from a few atom-mechanisms.

Switchable Whitening for Deep Representation Learning

1 code implementation ICCV 2019 Xingang Pan, Xiaohang Zhan, Jianping Shi, Xiaoou Tang, Ping Luo

Unlike existing works that design normalization techniques for specific tasks, we propose Switchable Whitening (SW), which provides a general form unifying different whitening methods as well as standardization methods.

Domain Adaptation Image Classification +3

SSN: Learning Sparse Switchable Normalization via SparsestMax

1 code implementation CVPR 2019 Wenqi Shao, Tianjian Meng, Jingyu Li, Ruimao Zhang, Yudian Li, Xiaogang Wang, Ping Luo

Unlike $\ell_1$ and $\ell_0$ constraints that impose difficulties in optimization, we turn this constrained optimization problem into feed-forward computation by proposing SparsestMax, which is a sparse version of softmax.

FaceFeat-GAN: a Two-Stage Approach for Identity-Preserving Face Synthesis

no code implementations4 Dec 2018 Yujun Shen, Bolei Zhou, Ping Luo, Xiaoou Tang

In the second stage, they compete in the image domain to render photo-realistic images that contain high diversity but preserve identity.

Face Generation

Kalman Normalization: Normalizing Internal Representations Across Network Layers

no code implementations NeurIPS 2018 Guangrun Wang, Jiefeng Peng, Ping Luo, Xinjiang Wang, Liang Lin

In this paper, we present a novel normalization method, called Kalman Normalization (KN), for improving and accelerating the training of DNNs, particularly under the context of micro-batches.

object-detection Object Detection

Do Normalization Layers in a Deep ConvNet Really Need to Be Distinct?

no code implementations19 Nov 2018 Ping Luo, Zhanglin Peng, Jiamin Ren, Ruimao Zhang

Our results suggest that (1) using distinct normalizers improves both learning and generalization of a ConvNet; (2) the choices of normalizers are more related to depth and batch size, but less relevant to parameter initialization, learning rate decay, and solver; (3) different tasks and datasets have different behaviors when learning to select normalizers.

Towards Understanding Regularization in Batch Normalization

1 code implementation ICLR 2019 Ping Luo, Xinjiang Wang, Wenqi Shao, Zhanglin Peng

Batch Normalization (BN) improves both convergence and generalization in training neural networks.

Hierarchical Neural Network for Extracting Knowledgeable Snippets and Documents

no code implementations22 Aug 2018 Ganbin Zhou, Rongyu Cao, Xiang Ao, Ping Luo, Fen Lin, Leyu Lin, Qing He

Additionally, a "low-level sharing, high-level splitting" structure of CNN is designed to handle the documents from different content domains.

Temporal Sequence Distillation: Towards Few-Frame Action Recognition in Videos

no code implementations15 Aug 2018 Zhaoyang Zhang, Zhanghui Kuang, Ping Luo, Litong Feng, Wei zhang

Secondly, TSD significantly reduces the computations to run video action recognition with compressed frames on the cloud, while maintaining high recognition accuracies.

Action Recognition In Videos Temporal Action Localization

Two at Once: Enhancing Learning and Generalization Capacities via IBN-Net

16 code implementations ECCV 2018 Xingang Pan, Ping Luo, Jianping Shi, Xiaoou Tang

IBN-Net carefully integrates Instance Normalization (IN) and Batch Normalization (BN) as building blocks, and can be wrapped into many advanced deep networks to improve their performances.

Domain Adaptation

Talking Face Generation by Adversarially Disentangled Audio-Visual Representation

1 code implementation20 Jul 2018 Hang Zhou, Yu Liu, Ziwei Liu, Ping Luo, Xiaogang Wang

Talking face generation aims to synthesize a sequence of face images that correspond to a clip of speech.

Lip Reading Retrieval +2

SCAN: Self-and-Collaborative Attention Network for Video Person Re-identification

no code implementations16 Jul 2018 Ruimao Zhang, Hongbin Sun, Jingyu Li, Yuying Ge, Liang Lin, Ping Luo, Xiaogang Wang

To address the above issues, we present a novel and practical deep architecture for video person re-identification termed Self-and-Collaborative Attention Network (SCAN).

Video-Based Person Re-Identification

Differentiable Learning-to-Normalize via Switchable Normalization

3 code implementations ICLR 2019 Ping Luo, Jiamin Ren, Zhanglin Peng, Ruimao Zhang, Jingyu Li

We hope SN will help ease the usage and understand the normalization techniques in deep learning.