Search Results for author:
Found 82 papers, 32 papers with code
All You Need in Knowledge Distillation Is a Tailored Coordinate System
Knowledge Distillation (KD) is essential in transferring dark knowledge from a large teacher to a small student network, such that the student can be much more efficient than the teacher but with comparable accuracy.
Quantization without Tears
Deep neural networks, while achieving remarkable success across diverse tasks, demand significant resources, including computation, GPU memory, bandwidth, storage, and energy.
Diffusion Product Quantization
In this work, we explore the quantization of diffusion models in extreme compression regimes to reduce model size while maintaining performance.
Minimal Interaction Edge Tuning: A New Paradigm for Visual Adaptation
The rapid scaling of large vision pretrained models makes fine-tuning tasks more and more difficult on edge devices with low computational resources.
Effectively Compress KV Heads for LLM
The advent of pre-trained large language models (LLMs) has revolutionized various natural language processing tasks.
Unified Low-rank Compression Framework for Click-through Rate Prediction
To address these challenges, we propose a unified low-rank decomposition framework for compressing CTR prediction models.
On Improving the Algorithm-, Model-, and Data- Efficiency of Self-Supervised Learning
In this paper, we propose an efficient single-branch SSL method based on non-parametric instance discrimination, aiming to improve the algorithm, model, and data efficiency of SSL.
Dense Vision Transformer Compression with Few Samples
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.
DiffuLT: How to Make Diffusion Model Useful for Long-tail Recognition
This paper proposes a new pipeline for long-tail (LT) recognition.
Low-rank Attention Side-Tuning for Parameter-Efficient Fine-Tuning
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
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.
Rectify the Regression Bias in Long-Tailed Object Detection
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.
DTL: Disentangled Transfer Learning for Visual Recognition
When pre-trained models become rapidly larger, the cost of fine-tuning on downstream tasks steadily increases, too.
Multi-Label Self-Supervised Learning with Scene Images
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.
Quantized Feature Distillation for Network Quantization
QFD first trains a quantized (or binarized) representation as the teacher, then quantize the network using knowledge distillation (KD).
Coarse Is Better? A New Pipeline Towards Self-Supervised Learning with Uncurated Images
That is, coarse crops benefits scene images SSL.
Instance-based Max-margin for Practical Few-shot Recognition
With both the new pFSL setting and novel IbM2 method, this paper shows that practical few-shot learning is both viable and promising.
No One Left Behind: Improving the Worst Categories in Long-Tailed Learning
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
For 22% latency reduction, it surpasses previous methods by on average 7 percentage points on ImageNet-1k.
Synergistic Self-supervised and Quantization Learning
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.
Worst Case Matters for Few-Shot Recognition
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.
R2-D2: Repetitive Reprediction Deep Decipher for Semi-Supervised Deep Learning
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
It is easy to collect a dataset with noisy labels, but such noise makes networks overfit seriously and accuracies drop dramatically.
Practical Network Acceleration with Tiny Sets
Previous methods mainly adopt filter-level pruning to accelerate networks with scarce training samples.
ActionFormer: Localizing Moments of Actions with Transformers
Self-attention based Transformer models have demonstrated impressive results for image classification and object detection, and more recently for video understanding.
Ranked #2 on
audio-visual event localization
on UnAV-100
Training Vision Transformers with Only 2040 Images
Vision Transformers (ViTs) is emerging as an alternative to convolutional neural networks (CNNs) for visual recognition.
Compressing Models with Few Samples: Mimicking then Replacing
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
Recently, vision transformer (ViT) and its variants have achieved promising performances in various computer vision tasks.
Fine-Grained Image Analysis with Deep Learning: A Survey
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.
Webly Supervised Fine-Grained Recognition: Benchmark Datasets and An Approach
Learning from the web can ease the extreme dependence of deep learning on large-scale manually labeled datasets.
Residual Attention: A Simple but Effective Method for Multi-Label Recognition
Multi-label image recognition is a challenging computer vision task of practical use.
Ranked #1 on
Multi-Label Image Classification
on VOC2007
Multi-Label Image Classification
Multi-Label Image Recognition
Weakly Supervised Foreground Learning for Weakly Supervised Localization and Detection
Modern deep learning models require large amounts of accurately annotated data, which is often difficult to satisfy.
A Random CNN Sees Objects: One Inductive Bias of CNN and Its Applications
That is, a CNN has an inductive bias to naturally focus on objects, named as Tobias ("The object is at sight") in this paper.
Salvage of Supervision in Weakly Supervised Object Detection
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.
Bag of Tricks for Long-Tailed Visual Recognition with Deep Convolutional Neural Networks
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).
Friends and Foes in Learning from Noisy Labels
Learning from examples with noisy labels has attracted increasing attention recently.
Distilling Virtual Examples for Long-tailed Recognition
We tackle the long-tailed visual recognition problem from the knowledge distillation perspective by proposing a Distill the Virtual Examples (DiVE) method.
Ranked #22 on
Long-tail Learning
on iNaturalist 2018
Rethinking Self-Supervised Learning: Small is Beautiful
Self-supervised learning (SSL), in particular contrastive learning, has made great progress in recent years.
Mixup Without Hesitation
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.
Distilling Knowledge by Mimicking Features
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)
Learning Expectation of Label Distribution for Facial Age and Attractiveness Estimation
The effectiveness of our approach has been demonstrated on both facial age and attractiveness estimation tasks.
Ranked #1 on
Attractiveness Estimation
on CFD
Rethinking the Route Towards Weakly Supervised Object Localization
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)
Neural Network Pruning with Residual-Connections and Limited-Data
Knowledge distillation is an effective approach to compensate for the weakness of limited data.
Neural Random Subspace
The random subspace method, known as the pillar of random forests, is good at making precise and robust predictions.
Repetitive Reprediction Deep Decipher for Semi-Supervised Learning
In this paper, we propose a principled end-to-end framework named deep decipher (D2) for SSL.
Deep Learning for Fine-Grained Image Analysis: A Survey
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.
Towards Real-Time Action Recognition on Mobile Devices Using Deep Models
We show that pre-trained weights on ImageNet improve the accuracy under the real-time action recognition setting.
Probabilistic End-to-end Noise Correction for Learning with Noisy Labels
Deep learning has achieved excellent performance in various computer vision tasks, but requires a lot of training examples with clean labels.
Ranked #25 on
Image Classification
on Clothing1M
(using extra training data)
When Semi-Supervised Learning Meets Transfer Learning: Training Strategies, Models and Datasets
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).
Coarse-to-fine: A RNN-based hierarchical attention model for vehicle re-identification
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.
Age Estimation Using Expectation of Label Distribution Learning
Age estimation performance has been greatly improved by using convolutional neural network.
AutoPruner: An End-to-End Trainable Filter Pruning Method for Efficient Deep Model Inference
Previous filter pruning algorithms regard channel pruning and model fine-tuning as two independent steps.
Piecewise classifier mappings: Learning fine-grained learners for novel categories with few examples
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.
Vortex Pooling: Improving Context Representation in Semantic Segmentation
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.
Learning Effective Binary Visual Representations with Deep Networks
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.
Action Recognition with Coarse-to-Fine Deep Feature Integration and Asynchronous Fusion
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.
Code Attention: Translating Code to Comments by Exploiting Domain Features
Appropriate comments of code snippets provide insight for code functionality, which are helpful for program comprehension.
ThiNet: A Filter Level Pruning Method for Deep Neural Network Compression
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.
Sunrise or Sunset: Selective Comparison Learning for Subtle Attribute Recognition
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.
Adaptive Feeding: Achieving Fast and Accurate Detections by Adaptively Combining Object Detectors
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.
Unsupervised Object Discovery and Co-Localization by Deep Descriptor Transforming
Reusable model design becomes desirable with the rapid expansion of computer vision and machine learning applications.
Ranked #11 on
Single-object discovery
on COCO_20k
An Entropy-based Pruning Method for CNN Compression
Experiments on the ILSVRC-12 benchmark demonstrate the effectiveness of our method.
Deep Descriptor Transforming for Image Co-Localization
Reusable model design becomes desirable with the rapid expansion of machine learning applications.
Learning Correspondence Structures for Person Re-identification
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.
Deep Label Distribution Learning with Label Ambiguity
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.
Ranked #1 on
Head Pose Estimation
on BJUT-3D
A Tube-and-Droplet-based Approach for Representing and Analyzing Motion Trajectories
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.
Dense CNN Learning with Equivalent Mappings
Large receptive field and dense prediction are both important for achieving high accuracy in pixel labeling tasks such as semantic segmentation.
Mask-CNN: Localizing Parts and Selecting Descriptors for Fine-Grained Image Recognition
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.
Selective Convolutional Descriptor Aggregation for Fine-Grained Image Retrieval
Moreover, on general image retrieval datasets, SCDA achieves comparable retrieval results with state-of-the-art general image retrieval approaches.
Minimal Gated Unit for Recurrent Neural Networks
Recently recurrent neural networks (RNN) has been very successful in handling sequence data.
Structured Learning of Binary Codes with Column Generation
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.
A diffusion and clustering-based approach for finding coherent motions and understanding crowd scenes
These semantic regions can be used to recognize pre-defined activities in crowd scenes.
Person Re-identification with Correspondence Structure Learning
This paper addresses the problem of handling spatial misalignments due to camera-view changes or human-pose variations in person re-identification.
Exploit Bounding Box Annotations for Multi-label Object Recognition
With strong labels, our framework is able to achieve state-of-the-art results in both datasets.
Ranked #17 on
Multi-Label Classification
on PASCAL VOC 2007
Deep Spatial Pyramid: The Devil is Once Again in the Details
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.
Weakly Supervised Fine-Grained Image Categorization
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.
Visual Recognition Using Directional Distribution Distance
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
In this paper, a new heat-map-based (HMB) algorithm is proposed for group activity recognition.
A new network-based algorithm for human activity recognition in video
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.
Optimizing Ranking Measures for Compact Binary Code Learning
Hashing has proven a valuable tool for large-scale information retrieval.
Towards Good Practices for Action Video Encoding
High dimensional representations such as VLAD or FV have shown excellent accuracy in action recognition.
Compact Representation for Image Classification: To Choose or to Compress?
In spite of the popularity of various feature compression methods, this paper argues that feature selection is a better choice than feature compression.
