Search Results for author: Jianxin Wu

Found 75 papers, 31 papers with code

Dense Vision Transformer Compression with Few Samples

no code implementations27 Mar 2024 Hanxiao Zhang, Yifan Zhou, Guo-Hua Wang, Jianxin Wu

In particular, the issue of sparse compression exists in traditional CNN few-shot methods, which can only produce very few compressed models of different model sizes.

Model Compression

Low-rank Attention Side-Tuning for Parameter-Efficient Fine-Tuning

no code implementations6 Feb 2024 Ningyuan Tang, Minghao Fu, Ke Zhu, Jianxin Wu

Because learnable parameters from these methods are entangled with the pretrained model, gradients related to the frozen pretrained model's parameters have to be computed and stored during finetuning.

Reviving Undersampling for Long-Tailed Learning

1 code implementation30 Jan 2024 Hao Yu, Yingxiao Du, Jianxin Wu

In this paper, we aim to enhance the accuracy of the worst-performing categories and utilize the harmonic mean and geometric mean to assess the model's performance.

Few-Shot Learning

Rectify the Regression Bias in Long-Tailed Object Detection

no code implementations29 Jan 2024 Ke Zhu, Minghao Fu, Jie Shao, Tianyu Liu, Jianxin Wu

While existing methods fail to handle the regression bias, the class-specific regression head for rare classes is hypothesized to be the main cause of it in this paper.

Long-tailed Object Detection Object +3

DTL: Disentangled Transfer Learning for Visual Recognition

1 code implementation13 Dec 2023 Minghao Fu, Ke Zhu, Jianxin Wu

When pre-trained models become rapidly larger, the cost of fine-tuning on downstream tasks steadily increases, too.

Transfer Learning

Multi-Label Self-Supervised Learning with Scene Images

no code implementations ICCV 2023 Ke Zhu, Minghao Fu, Jianxin Wu

Self-supervised learning (SSL) methods targeting scene images have seen a rapid growth recently, and they mostly rely on either a dedicated dense matching mechanism or a costly unsupervised object discovery module.

Contrastive Learning Multi-Label Classification +2

Quantized Feature Distillation for Network Quantization

no code implementations20 Jul 2023 Ke Zhu, Yin-Yin He, Jianxin Wu

QFD first trains a quantized (or binarized) representation as the teacher, then quantize the network using knowledge distillation (KD).

Image Classification Image Segmentation +6

Instance-based Max-margin for Practical Few-shot Recognition

no code implementations27 May 2023 Minghao Fu, Ke Zhu, Jianxin Wu

With both the new pFSL setting and novel IbM2 method, this paper shows that practical few-shot learning is both viable and promising.

Few-Shot Learning

No One Left Behind: Improving the Worst Categories in Long-Tailed Learning

no code implementations CVPR 2023 Yingxiao Du, Jianxin Wu

The convention in long-tailed recognition is to manually split all categories into three subsets and report the average accuracy within each subset.

Practical Network Acceleration with Tiny Sets: Hypothesis, Theory, and Algorithm

1 code implementation2 Mar 2023 Guo-Hua Wang, Jianxin Wu

For 22% latency reduction, it surpasses previous methods by on average 7 percentage points on ImageNet-1k.

Synergistic Self-supervised and Quantization Learning

1 code implementation12 Jul 2022 Yun-Hao Cao, Peiqin Sun, Yechang Huang, Jianxin Wu, Shuchang Zhou

In this paper, we propose a method called synergistic self-supervised and quantization learning (SSQL) to pretrain quantization-friendly self-supervised models facilitating downstream deployment.

Quantization Self-Supervised Learning

Worst Case Matters for Few-Shot Recognition

1 code implementation13 Mar 2022 Minghao Fu, Yun-Hao Cao, Jianxin Wu

Few-shot recognition learns a recognition model with very few (e. g., 1 or 5) images per category, and current few-shot learning methods focus on improving the average accuracy over many episodes.

Few-Shot Image Classification Few-Shot Learning

R2-D2: Repetitive Reprediction Deep Decipher for Semi-Supervised Deep Learning

no code implementations18 Feb 2022 Guo-Hua Wang, Jianxin Wu

However, they lack theoretical support and cannot explain why predictions are good candidates for pseudo-labels in the deep learning paradigm.

PENCIL: Deep Learning with Noisy Labels

no code implementations17 Feb 2022 Kun Yi, Guo-Hua Wang, Jianxin Wu

It is easy to collect a dataset with noisy labels, but such noise makes networks overfit seriously and accuracies drop dramatically.

Learning with noisy labels Multi-Label Classification

Practical Network Acceleration with Tiny Sets

1 code implementation CVPR 2023 Guo-Hua Wang, Jianxin Wu

Previous methods mainly adopt filter-level pruning to accelerate networks with scarce training samples.

ActionFormer: Localizing Moments of Actions with Transformers

1 code implementation16 Feb 2022 Chenlin Zhang, Jianxin Wu, Yin Li

Self-attention based Transformer models have demonstrated impressive results for image classification and object detection, and more recently for video understanding.

Action Recognition audio-visual event localization +3

Training Vision Transformers with Only 2040 Images

2 code implementations26 Jan 2022 Yun-Hao Cao, Hao Yu, Jianxin Wu

Vision Transformers (ViTs) is emerging as an alternative to convolutional neural networks (CNNs) for visual recognition.

Inductive Bias

Compressing Models with Few Samples: Mimicking then Replacing

1 code implementation CVPR 2022 Huanyu Wang, Junjie Liu, Xin Ma, Yang Yong, Zhenhua Chai, Jianxin Wu

Hence, previous methods optimize the compressed model layer-by-layer and try to make every layer have the same outputs as the corresponding layer in the teacher model, which is cumbersome.

A Unified Pruning Framework for Vision Transformers

1 code implementation30 Nov 2021 Hao Yu, Jianxin Wu

Recently, vision transformer (ViT) and its variants have achieved promising performances in various computer vision tasks.

Model Compression object-detection +1

Fine-Grained Image Analysis with Deep Learning: A Survey

no code implementations11 Nov 2021 Xiu-Shen Wei, Yi-Zhe Song, Oisin Mac Aodha, Jianxin Wu, Yuxin Peng, Jinhui Tang, Jian Yang, Serge Belongie

Fine-grained image analysis (FGIA) is a longstanding and fundamental problem in computer vision and pattern recognition, and underpins a diverse set of real-world applications.

Fine-Grained Image Recognition Image Retrieval +1

Weakly Supervised Foreground Learning for Weakly Supervised Localization and Detection

no code implementations3 Aug 2021 Chen-Lin Zhang, Yin Li, Jianxin Wu

Modern deep learning models require large amounts of accurately annotated data, which is often difficult to satisfy.

Weakly-Supervised Object Localization

A Random CNN Sees Objects: One Inductive Bias of CNN and Its Applications

1 code implementation17 Jun 2021 Yun-Hao Cao, Jianxin Wu

That is, a CNN has an inductive bias to naturally focus on objects, named as Tobias ("The object is at sight") in this paper.

Inductive Bias Object +3

Salvage of Supervision in Weakly Supervised Object Detection

no code implementations CVPR 2022 Lin Sui, Chen-Lin Zhang, Jianxin Wu

However, the lack of bounding-box supervision makes its accuracy much lower than fully supervised object detection (FSOD), and currently modern FSOD techniques cannot be applied to WSOD.

Object object-detection +2

Bag of Tricks for Long-Tailed Visual Recognition with Deep Convolutional Neural Networks

2 code implementations Association for the Advancement of Artificial Intelligence 2021 Yongshun Zhang, Xiu-Shen Wei, Boyan Zhou, Jianxin Wu

In recent years, visual recognition on challenging long-tailed distributions, where classes often exhibit extremely imbalanced frequencies, has made great progress mostly based on various complex paradigms (e. g., meta learning).

Data Augmentation Meta-Learning

Distilling Virtual Examples for Long-tailed Recognition

1 code implementation ICCV 2021 Yin-Yin He, Jianxin Wu, Xiu-Shen Wei

We tackle the long-tailed visual recognition problem from the knowledge distillation perspective by proposing a Distill the Virtual Examples (DiVE) method.

Knowledge Distillation Long-tail Learning

Friends and Foes in Learning from Noisy Labels

no code implementations28 Mar 2021 Yifan Zhou, Yifan Ge, Jianxin Wu

Learning from examples with noisy labels has attracted increasing attention recently.

Self-Supervised Learning valid

Rethinking Self-Supervised Learning: Small is Beautiful

1 code implementation25 Mar 2021 Yun-Hao Cao, Jianxin Wu

Self-supervised learning (SSL), in particular contrastive learning, has made great progress in recent years.

Contrastive Learning Self-Supervised Learning

Mixup Without Hesitation

1 code implementation12 Jan 2021 Hao Yu, Huanyu Wang, Jianxin Wu

In this paper, we find that mixup constantly explores the representation space, and inspired by the exploration-exploitation dilemma in reinforcement learning, we propose mixup Without hesitation (mWh), a concise, effective, and easy-to-use training algorithm.

Data Augmentation Image Classification +2

Distilling Knowledge by Mimicking Features

3 code implementations3 Nov 2020 Guo-Hua Wang, Yifan Ge, Jianxin Wu

We argue that the teacher should give more freedom to the student feature's magnitude, and let the student pay more attention on mimicking the feature direction.

 Ranked #1 on Knowledge Distillation on MS COCO (mAP metric)

Knowledge Distillation object-detection +1

Rethinking the Route Towards Weakly Supervised Object Localization

1 code implementation CVPR 2020 Chen-Lin Zhang, Yun-Hao Cao, Jianxin Wu

Weakly supervised object localization (WSOL) aims to localize objects with only image-level labels.

Ranked #2 on Weakly-Supervised Object Localization on CUB-200-2011 (Top-1 Localization Accuracy metric)

General Classification Object +1

Neural Random Subspace

1 code implementation18 Nov 2019 Yun-Hao Cao, Jianxin Wu, Hanchen Wang, Joan Lasenby

The random subspace method, known as the pillar of random forests, is good at making precise and robust predictions.

Representation Learning

Deep Learning for Fine-Grained Image Analysis: A Survey

1 code implementation6 Jul 2019 Xiu-Shen Wei, Jianxin Wu, Quan Cui

Among various research areas of CV, fine-grained image analysis (FGIA) is a longstanding and fundamental problem, and has become ubiquitous in diverse real-world applications.

Fine-Grained Image Recognition Image Generation +2

Towards Real-Time Action Recognition on Mobile Devices Using Deep Models

no code implementations17 Jun 2019 Chen-Lin Zhang, Xin-Xin Liu, Jianxin Wu

We show that pre-trained weights on ImageNet improve the accuracy under the real-time action recognition setting.

Action Recognition Hand Gesture Recognition +1

Probabilistic End-to-end Noise Correction for Learning with Noisy Labels

3 code implementations CVPR 2019 Kun Yi, Jianxin Wu

Deep learning has achieved excellent performance in various computer vision tasks, but requires a lot of training examples with clean labels.

Ranked #21 on Image Classification on Clothing1M (using extra training data)

Learning with noisy labels

When Semi-Supervised Learning Meets Transfer Learning: Training Strategies, Models and Datasets

no code implementations13 Dec 2018 Hong-Yu Zhou, Avital Oliver, Jianxin Wu, Yefeng Zheng

While practitioners have had an intuitive understanding of these observations, we do a comprehensive emperical analysis and demonstrate that: (1) the gains from SSL techniques over a fully-supervised baseline are smaller when trained from a pre-trained model than when trained from random initialization, (2) when the domain of the source data used to train the pre-trained model differs significantly from the domain of the target task, the gains from SSL are significantly higher and (3) some SSL methods are able to advance fully-supervised baselines (like Pseudo-Label).

Pseudo Label Transfer Learning

Coarse-to-fine: A RNN-based hierarchical attention model for vehicle re-identification

no code implementations11 Dec 2018 Xiu-Shen Wei, Chen-Lin Zhang, Lingqiao Liu, Chunhua Shen, Jianxin Wu

Inspired by the coarse-to-fine hierarchical process, we propose an end-to-end RNN-based Hierarchical Attention (RNN-HA) classification model for vehicle re-identification.

Vehicle Re-Identification

Age Estimation Using Expectation of Label Distribution Learning

1 code implementation13 Jul 2018 Bin-Bin Gao, Hong-Yu Zhou, Jianxin Wu, Xin Geng

Age estimation performance has been greatly improved by using convolutional neural network.

Age Estimation Face Recognition +1

AutoPruner: An End-to-End Trainable Filter Pruning Method for Efficient Deep Model Inference

no code implementations23 May 2018 Jian-Hao Luo, Jianxin Wu

Previous filter pruning algorithms regard channel pruning and model fine-tuning as two independent steps.

Binarization

Piecewise classifier mappings: Learning fine-grained learners for novel categories with few examples

1 code implementation11 May 2018 Xiu-Shen Wei, Peng Wang, Lingqiao Liu, Chunhua Shen, Jianxin Wu

To solve this problem, we propose an end-to-end trainable deep network which is inspired by the state-of-the-art fine-grained recognition model and is tailored for the FSFG task.

Few-Shot Learning Fine-Grained Image Recognition

Vortex Pooling: Improving Context Representation in Semantic Segmentation

no code implementations17 Apr 2018 Chen-Wei Xie, Hong-Yu Zhou, Jianxin Wu

To be specific, our approach outperforms the previous state-of-the-art model named DeepLab v3 by 1. 5% on the PASCAL VOC 2012 val set and 0. 6% on the test set by replacing the Atrous Spatial Pyramid Pooling (ASPP) module in DeepLab v3 with the proposed Vortex Pooling.

Semantic Segmentation

Learning Effective Binary Visual Representations with Deep Networks

no code implementations8 Mar 2018 Jianxin Wu, Jian-Hao Luo

Although traditionally binary visual representations are mainly designed to reduce computational and storage costs in the image retrieval research, this paper argues that binary visual representations can be applied to large scale recognition and detection problems in addition to hashing in retrieval.

General Classification Image Retrieval +3

Action Recognition with Coarse-to-Fine Deep Feature Integration and Asynchronous Fusion

no code implementations20 Nov 2017 Weiyao Lin, Yang Mi, Jianxin Wu, Ke Lu, Hongkai Xiong

In this paper, we propose a novel deep-based framework for action recognition, which improves the recognition accuracy by: 1) deriving more precise features for representing actions, and 2) reducing the asynchrony between different information streams.

Action Recognition Temporal Action Localization

Code Attention: Translating Code to Comments by Exploiting Domain Features

2 code implementations22 Sep 2017 Wenhao Zheng, Hong-Yu Zhou, Ming Li, Jianxin Wu

Appropriate comments of code snippets provide insight for code functionality, which are helpful for program comprehension.

Comment Generation

Adaptive Feeding: Achieving Fast and Accurate Detections by Adaptively Combining Object Detectors

no code implementations ICCV 2017 Hong-Yu Zhou, Bin-Bin Gao, Jianxin Wu

In this paper, we propose Adaptive Feeding (AF) to combine a fast (but less accurate) detector and an accurate (but slow) detector, by adaptively determining whether an image is easy or hard and choosing an appropriate detector for it.

object-detection Object Detection

ThiNet: A Filter Level Pruning Method for Deep Neural Network Compression

no code implementations ICCV 2017 Jian-Hao Luo, Jianxin Wu, Weiyao Lin

Similar experiments with ResNet-50 reveal that even for a compact network, ThiNet can also reduce more than half of the parameters and FLOPs, at the cost of roughly 1$\%$ top-5 accuracy drop.

Neural Network Compression

Sunrise or Sunset: Selective Comparison Learning for Subtle Attribute Recognition

no code implementations20 Jul 2017 Hong-Yu Zhou, Bin-Bin Gao, Jianxin Wu

The difficulty of image recognition has gradually increased from general category recognition to fine-grained recognition and to the recognition of some subtle attributes such as temperature and geolocation.

Attribute

An Entropy-based Pruning Method for CNN Compression

no code implementations19 Jun 2017 Jian-Hao Luo, Jianxin Wu

Experiments on the ILSVRC-12 benchmark demonstrate the effectiveness of our method.

Deep Descriptor Transforming for Image Co-Localization

no code implementations8 May 2017 Xiu-Shen Wei, Chen-Lin Zhang, Yao Li, Chen-Wei Xie, Jianxin Wu, Chunhua Shen, Zhi-Hua Zhou

Reusable model design becomes desirable with the rapid expansion of machine learning applications.

Learning Correspondence Structures for Person Re-identification

no code implementations20 Mar 2017 Weiyao Lin, Yang shen, Junchi Yan, Mingliang Xu, Jianxin Wu, Jingdong Wang, Ke Lu

We first introduce a boosting-based approach to learn a correspondence structure which indicates the patch-wise matching probabilities between images from a target camera pair.

Patch Matching Person Re-Identification

Deep Label Distribution Learning with Label Ambiguity

2 code implementations6 Nov 2016 Bin-Bin Gao, Chao Xing, Chen-Wei Xie, Jianxin Wu, Xin Geng

However, it is difficult to collect sufficient training images with precise labels in some domains such as apparent age estimation, head pose estimation, multi-label classification and semantic segmentation.

Age Estimation Classification +4

A Tube-and-Droplet-based Approach for Representing and Analyzing Motion Trajectories

no code implementations10 Sep 2016 Weiyao Lin, Yang Zhou, Hongteng Xu, Junchi Yan, Mingliang Xu, Jianxin Wu, Zicheng Liu

Our approach first leverages the complete information from given trajectories to construct a thermal transfer field which provides a context-rich way to describe the global motion pattern in a scene.

3D Action Recognition Anomaly Detection +2

Dense CNN Learning with Equivalent Mappings

no code implementations24 May 2016 Jianxin Wu, Chen-Wei Xie, Jian-Hao Luo

Large receptive field and dense prediction are both important for achieving high accuracy in pixel labeling tasks such as semantic segmentation.

Age Estimation Image Categorization +2

Mask-CNN: Localizing Parts and Selecting Descriptors for Fine-Grained Image Recognition

no code implementations23 May 2016 Xiu-Shen Wei, Chen-Wei Xie, Jianxin Wu

Fine-grained image recognition is a challenging computer vision problem, due to the small inter-class variations caused by highly similar subordinate categories, and the large intra-class variations in poses, scales and rotations.

Fine-Grained Image Recognition

Selective Convolutional Descriptor Aggregation for Fine-Grained Image Retrieval

1 code implementation18 Apr 2016 Xiu-Shen Wei, Jian-Hao Luo, Jianxin Wu, Zhi-Hua Zhou

Moreover, on general image retrieval datasets, SCDA achieves comparable retrieval results with state-of-the-art general image retrieval approaches.

Image Retrieval Object Proposal Generation +1

Minimal Gated Unit for Recurrent Neural Networks

no code implementations31 Mar 2016 Guo-Bing Zhou, Jianxin Wu, Chen-Lin Zhang, Zhi-Hua Zhou

Recently recurrent neural networks (RNN) has been very successful in handling sequence data.

Structured Learning of Binary Codes with Column Generation

no code implementations22 Feb 2016 Guosheng Lin, Fayao Liu, Chunhua Shen, Jianxin Wu, Heng Tao Shen

Our column generation based method can be further generalized from the triplet loss to a general structured learning based framework that allows one to directly optimize multivariate performance measures.

Image Retrieval Information Retrieval +1

Person Re-identification with Correspondence Structure Learning

1 code implementation ICCV 2015 Yang Shen, Weiyao Lin, Junchi Yan, Mingliang Xu, Jianxin Wu, Jingdong Wang

This paper addresses the problem of handling spatial misalignments due to camera-view changes or human-pose variations in person re-identification.

Patch Matching Person Re-Identification

Deep Spatial Pyramid: The Devil is Once Again in the Details

no code implementations21 Apr 2015 Bin-Bin Gao, Xiu-Shen Wei, Jianxin Wu, Weiyao Lin

In this paper we show that by carefully making good choices for various detailed but important factors in a visual recognition framework using deep learning features, one can achieve a simple, efficient, yet highly accurate image classification system.

General Classification Image Classification

Weakly Supervised Fine-Grained Image Categorization

no code implementations20 Apr 2015 Yu Zhang, Xiu-Shen Wei, Jianxin Wu, Jianfei Cai, Jiangbo Lu, Viet-Anh Nguyen, Minh N. Do

Most existing works heavily rely on object / part detectors to build the correspondence between object parts by using object or object part annotations inside training images.

Fine-Grained Image Classification Image Categorization +1

Visual Recognition Using Directional Distribution Distance

no code implementations19 Apr 2015 Jianxin Wu, Bin-Bin Gao, Guoqing Liu

In computer vision, an entity such as an image or video is often represented as a set of instance vectors, which can be SIFT, motion, or deep learning feature vectors extracted from different parts of that entity.

A Heat-Map-based Algorithm for Recognizing Group Activities in Videos

no code implementations21 Feb 2015 Weiyao Lin, Hang Chu, Jianxin Wu, Bin Sheng, Zhenzhong Chen

In this paper, a new heat-map-based (HMB) algorithm is proposed for group activity recognition.

Group Activity Recognition

A new network-based algorithm for human activity recognition in video

no code implementations21 Feb 2015 Weiyao Lin, Yuanzhe Chen, Jianxin Wu, Hanli Wang, Bin Sheng, Hongxiang Li

Based on this network, we further model people in the scene as packages while human activities can be modeled as the process of package transmission in the network.

Activity Detection Activity Recognition In Videos +2

Compact Representation for Image Classification: To Choose or to Compress?

no code implementations CVPR 2014 Yu Zhang, Jianxin Wu, Jianfei Cai

In spite of the popularity of various feature compression methods, this paper argues that feature selection is a better choice than feature compression.

Classification Dimensionality Reduction +5

Cannot find the paper you are looking for? You can Submit a new open access paper.