Knowledge distillation, which involves extracting the "dark knowledge" from a teacher network to guide the learning of a student network, has emerged as an essential technique for model compression and transfer learning. Unlike previous works that focus on the accuracy of student network, here we study a little-explored but important question, i.e., knowledge distillation efficiency. Our goal is to achieve a performance comparable to conventional knowledge distillation with a lower computation cost during training. We show that the UNcertainty-aware mIXup (UNIX) can serve as a clean yet effective solution. The uncertainty sampling strategy is used to evaluate the informativeness of each training sample. Adaptive mixup is applied to uncertain samples to compact knowledge. We further show that the redundancy of conventional knowledge distillation lies in the excessive learning of easy samples. By combining uncertainty and mixup, our approach reduces the redundancy and makes better use of each query to the teacher network. We validate our approach on CIFAR100 and ImageNet. Notably, with only 79% computation cost, we outperform conventional knowledge distillation on CIFAR100 and achieve a comparable result on ImageNet.
Mixup is a popular data augmentation technique based on creating new samples by linear interpolation between two given data samples, to improve both the generalization and robustness of the trained model. Knowledge distillation (KD), on the other hand, is widely used for model compression and transfer learning, which involves using a larger network's implicit knowledge to guide the learning of a smaller network. At first glance, these two techniques seem very different, however, we found that "smoothness" is the connecting link between the two and is also a crucial attribute in understanding K…
Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Horizon Robotics（地平线机器人）
Unlike the conventional Knowledge Distillation (KD), Self-KD allows a network to learn knowledge from itself without any guidance from extra networks. This paper proposes to perform Self-KD from image Mixture (MixSKD), which integrates these two techniques into a unified framework. MixSKD mutually distills feature maps and probability distributions between the random pair of original images and their mixup images in a meaningful way. Therefore, it guides the network to learn cross-image knowledge by modelling supervisory signals from mixup images. Moreover, we construct a self-teacher network …
We study how to train a student deep neural network for visual recognition by distilling knowledge from a blackbox teacher model in a data-efficient manner. Progress on this problem can significantly reduce the dependence on large-scale datasets for learning high-performing visual recognition models. There are two major challenges. One is that the number of queries into the teacher model should be minimized to save computational and/or financial costs. The other is that the number of images used for the knowledge distillation should be small; otherwise, it violates our expectation of reducing …
Beyond Dark Knowledge: Mixup-Based Distillation for Reliable Predictions
ITS Lab, Institute of Computer Science, University of Tartu（塔尔图大学计算机科学学院ITS实验室）；LITIS, Université de Rouen（鲁昂大学LITIS实验室）；LITIS, INSA de Rouen（鲁昂国立应用科学学院LITIS实验室）
Present domain adaptation methods usually perform explicit representation alignment by simultaneously accessing the source data and target data. However, the source data are not always available due to the privacy preserving consideration or bandwidth limitation. Source-free domain adaptation aims to solve the above problem by performing domain adaptation without accessing the source data. The adaptation paradigm is receiving more and more attention in recent years, and multiple works have been proposed for unsupervised source-free domain adaptation. However, without utilizing any supervised s…
Deep neural networks are susceptible to adversarially crafted, small and imperceptible changes in the natural inputs. The most effective defense mechanism against these examples is adversarial training which constructs adversarial examples during training by iterative maximization of loss. The model is then trained to minimize the loss on these constructed examples. This min-max optimization requires more data, larger capacity models, and additional computing resources. It also degrades the standard generalization performance of a model. Can we achieve robustness more efficiently? In this work…
Continual Self-Supervised Learning with Masked Autoencoders in Remote Sensing
Faculty of Electrical Engineering and Computer Science, Technische Universität Berlin（电气工程与计算机科学学院，柏林技术大学）；BIFOLD - Berlin Institute for the Foundations of Learning and Data（柏林学习与数据基础研究所）
Mixed Sample Regularization (MSR), such as MixUp or CutMix, is a powerful data augmentation strategy to generalize convolutional neural networks. Previous empirical analysis has illustrated an orthogonal performance gain between MSR and conventional offline Knowledge Distillation (KD). To be more specific, student networks can be enhanced with the involvement of MSR in the training stage of sequential distillation. Yet, the interplay between MSR and online knowledge distillation, where an ensemble of peer students learn mutually from each other, remains unexplored. To bridge the gap, we make t…
College of Information Science and Engineering, Ocean University of China（中国海洋大学信息科学与工程学院）；School of Creative Technologies, University of Portsmouth（朴茨茅斯大学创意技术学院）
Knowledge distillation is a popular paradigm for learning portable neural networks by transferring the knowledge from a large model into a smaller one. Most existing approaches enhance the student model by utilizing the similarity information between the categories of instance level provided by the teacher model. However, these works ignore the similarity correlation between different instances that plays an important role in confidence prediction. To tackle this issue, we propose a novel method in this paper, called similarity transfer for knowledge distillation (STKD), which aims to fully ut…
Uncertainty-Aware Retinal Vessel Segmentation via Ensemble Distillation
University of Ibadan（伊巴迪安大学）；African Institute for Mathematical Sciences（非洲数学科学研究所）；University College London（伦敦大学学院）；Hertie Institute for Brain Health（海德研究所）
Uncertainty-Aware Knowledge Distillation for Compact and Efficient 6DoF Pose Estimation
SnT, University of Luxembourg（卢森堡大学SnT分校）；Cristal Lab, ENSI, University of Manouba（曼努巴大学ENSI的Cristal实验室）；Infinite Orbits, Toulouse（图卢兹的Infinite Orbits）
National University of Defense Technology（国防科技大学）；University of Oulu（奥卢大学）
Given a model well-trained with a large-scale base dataset, Few-Shot Class-Incremental Learning (FSCIL) aims at incrementally learning novel classes from a few labeled samples by avoiding overfitting, without catastrophically forgetting all encountered classes previously. Currently, semi-supervised learning technique that harnesses freely-available unlabeled data to compensate for limited labeled data can boost the performance in numerous vision tasks, which heuristically can be applied to tackle issues in FSCIL, i.e., the Semi-supervised FSCIL (Semi-FSCIL). So far, very limited work focuses o…
Generalization Performance of Deep Learning models trained using Empirical Risk Minimization can be improved significantly by using Data Augmentation strategies such as simple transformations, or using Mixed Samples. We attempt to empirically analyze the impact of such strategies on the transfer of generalization between teacher and student models in a distillation setup. We observe that if a teacher is trained using any of the mixed sample augmentation strategies, such as MixUp or CutMix, the student model distilled from it is impaired in its generalization capabilities. We hypothesize that s…
Rethinking Data Distillation: Do Not Overlook Calibration
Texas A&M University（德州农工大学）；Zhejiang University（浙江大学）；Certik；University of Maryland, College Park（马里兰大学帕克分校）；Purdue University（普渡大学）；North Carolina State University（北卡罗来纳州立大学）
Hong Kong University of Science and Technology（香港科技大学）
When recognizing emotions, subtle nuances in displays of emotion generate ambiguity or uncertainty in emotion perception. Emotion uncertainty has been previously interpreted as inter-rater disagreement among multiple annotators. In this paper, we consider a more common and challenging scenario: modeling emotion uncertainty when only single emotion labels are available. From a Bayesian perspective, we propose to use deep ensembles to capture uncertainty for multiple emotion descriptors, i.e., action units, discrete expression labels and continuous descriptors. We further apply iterative self-di…
Object re-identification (re-id) aims to identify a specific object across times or camera views, with the person re-id and vehicle re-id as the most widely studied applications. Re-id is challenging because of the variations in viewpoints, (human) poses, and occlusions. Multi-shots of the same object can cover diverse viewpoints/poses and thus provide more comprehensive information. In this paper, we propose exploiting the multi-shots of the same identity to guide the feature learning of each individual image. Specifically, we design an Uncertainty-aware Multi-shot Teacher-Student (UMTS) Netw…
Uncertainty-Aware Dual-Student Knowledge Distillation for Efficient Image Classification
Department of Electrical Engineering Indian Institute of Technology Bombay 21D070002（电子工程系 印度理工学院班加罗尔 21D070002）；Department of Electrical Engineering Indian Institute of Technology Bombay 210010010（电子工程系 印度理工学院班加罗尔 210010010）；Department of Electrical Engineering Indian Institute of Technology Bombay 210020023（电子工程系 印度理工学院班加罗尔 210020023）
Applied Artificial Intelligence Institute (AI), Deakin University（迪肯大学应用人工智能研究所）
Knowledge distillation (KD) is an efficient approach to transfer the knowledge from a large "teacher" network to a smaller "student" network. Traditional KD methods require lots of labeled training samples and a white-box teacher (parameters are accessible) to train a good student. However, these resources are not always available in real-world applications. The distillation process often happens at an external party side where we do not have access to much data, and the teacher does not disclose its parameters due to security and privacy concerns. To overcome these challenges, we propose a bl…
We propose an efficient way to output better calibrated uncertainty scores from neural networks. The Distilled Dropout Network (DDN) makes standard (non-Bayesian) neural networks more introspective by adding a new training loss which prevents them from being overconfident. Our method is more efficient than Bayesian neural networks or model ensembles which, despite providing more reliable uncertainty scores, are more cumbersome to train and slower to test. We evaluate DDN on the the task of image classification on the CIFAR-10 dataset and show that our calibration results are competitive even w…
Indian Institute of Technology, Jodhpur（焦特布尔印度理工学院）
Knowledge distillation is a common technique for improving the performance of a shallow student network by transferring information from a teacher network, which in general, is comparatively large and deep. These teacher networks are pre-trained and often uncalibrated, as no calibration technique is applied to the teacher model while training. Calibration of a network measures the probability of correctness for any of its predictions, which is critical in high-risk domains. In this paper, we study how to obtain a calibrated student from an uncalibrated teacher. Our approach relies on the fusio…
Current deep learning models often suffer from catastrophic forgetting of old knowledge when continually learning new knowledge. Existing strategies to alleviate this issue often fix the trade-off between keeping old knowledge (stability) and learning new knowledge (plasticity). However, the stability-plasticity trade-off during continual learning may need to be dynamically changed for better model performance. In this paper, we propose two novel ways to adaptively balance model stability and plasticity. The first one is to adaptively integrate multiple levels of old knowledge and transfer it …
Bayesian models have many desirable properties, most notable is their ability to generalize from limited data and to properly estimate the uncertainty in their predictions. However, these benefits come at a steep computational cost as Bayesian inference, in most cases, is computationally intractable. One popular approach to alleviate this problem is using a Monte-Carlo estimation with an ensemble of models sampled from the posterior. However, this approach still comes at a significant computational cost, as one needs to store and run multiple models at test time. In this work, we investigate h…
University of Science and Technology of China（中国科学技术大学）；Megvii Inc.（旷视科技）；Tongji University（同济大学）
Knowledge distillation(KD) aims to improve the performance of a student network by mimicing the knowledge from a powerful teacher network. Existing methods focus on studying what knowledge should be transferred and treat all samples equally during training. This paper introduces the adaptive sample weighting to KD. We discover that previous effective hard mining methods are not appropriate for distillation. Furthermore, we propose Prime-Aware Adaptive Distillation (PAD) by the incorporation of uncertainty learning. PAD perceives the prime samples in distillation and then emphasizes their effec…
Uncertainty-Aware Multi-Expert Knowledge Distillation for Imbalanced Disease Grading
School of Public Health, Zhejiang University（浙江大学公共卫生学院）；College of Computer Science and Technology, Zhejiang University（浙江大学计算机科学与技术学院）；Liangzhu Laboratory and WeDoctor Cloud（良渚实验室和WeDoctor云）；State Key Laboratory of Transvascular Implantation Devices, The Second Affiliated Hospital Zhejiang University School of Medicine（浙江大学医学院第二附属医院心血管植入器械国家重点实验室）；Zhejiang Key Laboratory of Medical Imaging Artificial Intelligence（浙江省医学影像人工智能重点实验室）；School of Electronic Information and Communications, Huazhong University of Science and Technology（华中科技大学电子信息与通信学院）
Steklov Mathematical Institute（斯捷克洛夫数学研究所）；Neuromation OU；Samsung AI Center（三星人工智能中心）
Despite recent advances in computer vision based on various convolutional architectures, video understanding remains an important challenge. In this work, we present and discuss a top solution for the large-scale video classification (labeling) problem introduced as a Kaggle competition based on the YouTube-8M dataset. We show and compare different approaches to preprocessing, data augmentation, model architectures, and model combination. Our final model is based on a large ensemble of video- and frame-level models but fits into rather limiting hardware constraints. We apply an approach based …
Test-Time Mixup Augmentation for Data and Class-Specific Uncertainty Estimation in Deep Learning Image Classification
School of Electrical Engineering（电气工程学院）；Korea Advanced Institute of Science and Technology（韩国科学技术院）；Department of Software Convergence（软件融合系）；College of Interdisciplinary Studies for Emerging Industries（新兴产业跨学科研究学院）；Seoul Women’s University（首尔女子大学）
We introduce a simple yet effective distillation framework that is able to boost the vanilla ResNet-50 to 80%+ Top-1 accuracy on ImageNet without tricks. We construct such a framework through analyzing the problems in the existing classification system and simplify the base method ensemble knowledge distillation via discriminators by: (1) adopting the similarity loss and discriminator only on the final outputs and (2) using the average of softmax probabilities from all teacher ensembles as the stronger supervision. Intriguingly, three novel perspectives are presented for distillation: (1) weig…
One of the major limitations of deep learning models is that they face catastrophic forgetting in an incremental learning scenario. There have been several approaches proposed to tackle the problem of incremental learning. Most of these methods are based on knowledge distillation and do not adequately utilize the information provided by older task models, such as uncertainty estimation in predictions. The predictive uncertainty provides the distributional information can be applied to mitigate catastrophic forgetting in a deep learning framework. In the proposed work, we consider a Bayesian fo…
University of California, Merced（加州大学默塞德分校）；eBay Inc.（eBay公司）；NEC Labs America（美国NEC实验室）；Google Research（谷歌研究院）
Obtaining object response maps is one important step to achieve weakly-supervised semantic segmentation using image-level labels. However, existing methods rely on the classification task, which could result in a response map only attending on discriminative object regions as the network does not need to see the entire object for optimizing the classification loss. To tackle this issue, we propose a principled and end-to-end train-able framework to allow the network to pay attention to other parts of the object, while producing a more complete and uniform response map. Specifically, we introdu…
University of Oxford（牛津大学）；Johns Hopkins University（约翰斯·霍普金斯大学）；The University of Hong Kong（香港大学）；ByteDance Inc.（字节跳动公司）
Modern deep networks can be better generalized when trained with noisy samples and regularization techniques. Mixup and CutMix have been proven to be effective for data augmentation to help avoid overfitting. Previous Mixup-based methods linearly combine images and labels to generate additional training data. However, this is problematic if the object does not occupy the whole image as we demonstrate in Figure 1. Correctly assigning the label weights is hard even for human beings and there is no clear criterion to measure it. To tackle this problem, in this paper, we propose LUMix, which model…
Negative sampling is a limiting factor w.r.t. the generalization of metric-learned neural networks. We show that uniform negative sampling provides little information about the class boundaries and thus propose three novel techniques for efficient negative sampling: drawing negative samples from (1) the top-$k$ most semantically similar classes, (2) the top-$k$ most semantically similar samples and (3) interpolating between contrastive latent representations to create pseudo negatives. Our experiments on CIFAR-10, CIFAR-100 and Tiny-ImageNet-200 show that our proposed \textit{Semantically Cond…
University of Oxford（牛津大学）；Meta AI；Five AI Ltd.（Five AI有限公司）
We show that the effectiveness of the well celebrated Mixup [Zhang et al., 2018] can be further improved if instead of using it as the sole learning objective, it is utilized as an additional regularizer to the standard cross-entropy loss. This simple change not only provides much improved accuracy but also significantly improves the quality of the predictive uncertainty estimation of Mixup in most cases under various forms of covariate shifts and out-of-distribution detection experiments. In fact, we observe that Mixup yields much degraded performance on detecting out-of-distribution samples …
University of Padova（帕多瓦大学）；University of Catania（卡塔尼亚大学）
Human capability to anticipate near future from visual observations and non-verbal cues is essential for developing intelligent systems that need to interact with people. Several research areas, such as human-robot interaction (HRI), assisted living or autonomous driving need to foresee future events to avoid crashes or help people. Egocentric scenarios are classic examples where action anticipation is applied due to their numerous applications. Such challenging task demands to capture and model domain's hidden structure to reduce prediction uncertainty. Since multiple actions may equally occu…
SUMix: Mixup with Semantic and Uncertain Information
Chongqing Technology and Business University（重庆工商大学）；National Research base of Intelligent Manufacturing Service（智能制造服务国家研究基地）；Chongqing Micro-Vein Intelligent Technology Co.（重庆微脉智能科技有限公司）；Telecom SudParis（巴黎电信学院）；Institut Polytechnique de Paris（巴黎综合理工学院）；Chongqing University of Posts and Telecommunications（重庆邮电大学）
McGill University（麦吉尔大学）；Centre for Intelligent Machines (CIM)（智能机器中心 (CIM)）；Department of Electrical and Computer Engineering（电气与计算机工程系）
Segmentation of enhancing tumours or lesions from MRI is important for detecting new disease activity in many clinical contexts. However, accurate segmentation requires the inclusion of medical images (e.g., T1 post contrast MRI) acquired after injecting patients with a contrast agent (e.g., Gadolinium), a process no longer thought to be safe. Although a number of modality-agnostic segmentation networks have been developed over the past few years, they have been met with limited success in the context of enhancing pathology segmentation. In this work, we present HAD-Net, a novel offline advers…
School of Computing Sciences（计算科学学院）；University of East Anglia（东英吉利大学）
We propose SUB-Depth, a universal multi-task training framework for self-supervised monocular depth estimation (SDE). Depth models trained with SUB-Depth outperform the same models trained in a standard single-task SDE framework. By introducing an additional self-distillation task into a standard SDE training framework, SUB-Depth trains a depth network, not only to predict the depth map for an image reconstruction task, but also to distill knowledge from a trained teacher network with unlabelled data. To take advantage of this multi-task setting, we propose homoscedastic uncertainty formulatio…
Imperial College London（帝国理工学院）；National University of Singapore（新加坡国立大学）；National Institute of Technology Tiruchirappalli（蒂鲁吉拉伯利国家理工学院）；The Chinese University of Hong Kong（香港中文大学）；Shun Hing Institute of Advanced Engineering（信兴高等工程研究所）
Purpose: In curriculum learning, the idea is to train on easier samples first and gradually increase the difficulty, while in self-paced learning, a pacing function defines the speed to adapt the training progress. While both methods heavily rely on the ability to score the difficulty of data samples, an optimal scoring function is still under exploration. Methodology: Distillation is a knowledge transfer approach where a teacher network guides a student network by feeding a sequence of random samples. We argue that guiding student networks with an efficient curriculum strategy can improve mod…
Harbin Institute of Technology（哈尔滨工业大学）；University of Trento（特伦托大学）；Hong Kong University of Science and Technology（香港科技大学）；INRIA（法国国家信息与自动化研究所）；SAP AI Research（思爱普人工智能研究院）；Fondazione Bruno Kessler（布鲁诺·凯斯勒基金会）
A fundamental and challenging problem in deep learning is catastrophic forgetting, i.e. the tendency of neural networks to fail to preserve the knowledge acquired from old tasks when learning new tasks. This problem has been widely investigated in the research community and several Incremental Learning (IL) approaches have been proposed in the past years. While earlier works in computer vision have mostly focused on image classification and object detection, more recently some IL approaches for semantic segmentation have been introduced. These previous works showed that, despite its simplicity…
Beihang University（北京航空航天大学）；Institute for Infocomm Research, A*STAR（新加坡科技研究局资讯通信研究院）
This work aims to estimate a high-quality depth map from a single RGB image. Due to the lack of depth clues, making full use of the long-range correlation and the local information is critical for accurate depth estimation. Towards this end, we introduce an uncertainty rectified cross-distillation between Transformer and convolutional neural network (CNN) to learn a unified depth estimator. Specifically, we use the depth estimates from the Transformer branch and the CNN branch as pseudo labels to teach each other. Meanwhile, we model the pixel-wise depth uncertainty to rectify the loss weights…
