In order to train robust deep learning models, large amounts of labelled data is required. However, in the absence of such large repositories of labelled data, unlabeled data can be exploited for the same. Semi-Supervised learning aims to utilize such unlabeled data for training classification models. Recent progress of self-training based approaches have shown promise in this area, which leads to this study where we utilize an ensemble approach for the same. A by-product of any semi-supervised approach may be loss of calibration of the trained model especially in scenarios where unlabeled data may contain out-of-distribution samples, which leads to this investigation on how to adapt to such effects. Our proposed algorithm carefully avoids common pitfalls in utilizing unlabeled data and leads to a more accurate and calibrated supervised model compared to vanilla self-training based student-teacher algorithms. We perform several experiments on the popular STL-10 database followed by an extensive analysis of our approach and study its effects on model accuracy and calibration.
It has been recently demonstrated that multi-generational self-distillation can improve generalization. Despite this intriguing observation, reasons for the enhancement remain poorly understood. In this paper, we first demonstrate experimentally that the improved performance of multi-generational self-distillation is in part associated with the increasing diversity in teacher predictions. With this in mind, we offer a new interpretation for teacher-student training as amortized MAP estimation, such that teacher predictions enable instance-specific regularization. Our framework allows us to the…
Department of Basic Science, University of Tokyo（东京大学基础科学系）
On-line learning of a hierarchical learning model is studied by a method from statistical mechanics. In our model a student of a simple perceptron learns from not a true teacher directly, but ensemble teachers who learn from the true teacher with a perceptron learning rule. Since the true teacher and the ensemble teachers are expressed as non-monotonic perceptron and simple ones, respectively, the ensemble teachers go around the unlearnable true teacher with the distance between them fixed in an asymptotic steady state. The generalization performance of the student is shown to exceed that of t…
Traditional knowledge distillation transfers "dark knowledge" of a pre-trained teacher network to a student network, and ignores the knowledge in the training process of the teacher, which we call teacher's experience. However, in realistic educational scenarios, learning experience is often more important than learning results. In this work, we propose a novel knowledge distillation method by integrating the teacher's experience for knowledge transfer, named experience ensemble knowledge distillation (EEKD). We save a moderate number of intermediate models from the training process of the tea…
This paper proposes a novel knowledge distillation-based learning method to improve the classification performance of convolutional neural networks (CNNs) without a pre-trained teacher network, called exit-ensemble distillation. Our method exploits the multi-exit architecture that adds auxiliary classifiers (called exits) in the middle of a conventional CNN, through which early inference results can be obtained. The idea of our method is to train the network using the ensemble of the exits as the distillation target, which greatly improves the classification performance of the overall network.…
Graduate School of Convergence Science and Technology（融合科学与技术研究生院）；Seoul National University（首尔国立大学）
Knowledge Distillation (KD) aims to transfer knowledge in a teacher-student framework, by providing the predictions of the teacher network to the student network in the training stage to help the student network generalize better. It can use either a teacher with high capacity or {an} ensemble of multiple teachers. However, the latter is not convenient when one wants to use feature-map-based distillation methods. For a solution, this paper proposes a versatile and powerful training algorithm named FEature-level Ensemble for knowledge Distillation (FEED), which aims to transfer the ensemble kno…
In this paper, we present a novel approach, Momentum$^2$ Teacher, for student-teacher based self-supervised learning. The approach performs momentum update on both network weights and batch normalization (BN) statistics. The teacher's weight is a momentum update of the student, and the teacher's BN statistics is a momentum update of those in history. The Momentum$^2$ Teacher is simple and efficient. It can achieve the state of the art results (74.5\%) under ImageNet linear evaluation protocol using small-batch size(\eg, 128), without requiring large-batch training on special hardware like TPU …
Tsinghua University（清华大学）；Microsoft Research Asia（微软亚洲研究院）
Ensemble knowledge distillation can extract knowledge from multiple teacher models and encode it into a single student model. Many existing methods learn and distill the student model on labeled data only. However, the teacher models are usually learned on the same labeled data, and their predictions have high correlations with groudtruth labels. Thus, they cannot provide sufficient knowledge complementary to task labels for student teaching. Distilling on unseen unlabeled data has the potential to enhance the knowledge transfer from the teachers to the student. In this paper, we propose a uni…
Kobe City College of Technology（神户市立工业高等专门学校）；Kobe University（神户大学）；The University of Tokyo（东京大学）；RIKEN Brain Science Institute（理化学研究所脑科学研究所）
We analyze the generalization performance of a student in a model composed of nonlinear perceptrons: a true teacher, ensemble teachers, and the student. We calculate the generalization error of the student analytically or numerically using statistical mechanics in the framework of on-line learning. We treat two well-known learning rules: Hebbian learning and perceptron learning. As a result, it is proven that the nonlinear model shows qualitatively different behaviors from the linear model. Moreover, it is clarified that Hebbian learning and perceptron learning show qualitatively different beh…
Stanford University（斯坦福大学）；University of Toronto（多伦多大学）；Vector Institute（矢量研究所）
The exploration mechanism used by a Deep Reinforcement Learning (RL) agent plays a key role in determining its sample efficiency. Thus, improving over random exploration is crucial to solve long-horizon tasks with sparse rewards. We propose to leverage an ensemble of partial solutions as teachers that guide the agent's exploration with action suggestions throughout training. While the setup of learning with teachers has been previously studied, our proposed approach - Actor-Critic with Teacher Ensembles (AC-Teach) - is the first to work with an ensemble of suboptimal teachers that may solve on…
Knowledge distillation usually transfers the knowledge from a pre-trained cumbersome teacher network to a compact student network, which follows the classical teacher-teaching-student paradigm. Based on this paradigm, previous methods mostly focus on how to efficiently train a better student network for deployment. Different from the existing practices, in this paper, we propose a novel student-helping-teacher formula, Teacher Evolution via Self-Knowledge Distillation (TESKD), where the target teacher (for deployment) is learned with the help of multiple hierarchical students by sharing the st…
Recent advances in deep learning have facilitated the demand of neural models for real applications. In practice, these applications often need to be deployed with limited resources while keeping high accuracy. This paper touches the core of neural models in NLP, word embeddings, and presents a new embedding distillation framework that remarkably reduces the dimension of word embeddings without compromising accuracy. A novel distillation ensemble approach is also proposed that trains a high-efficient student model using multiple teacher models. In our approach, the teacher models play roles on…
Privacy-preserving deep learning is crucial for deploying deep neural network based solutions, especially when the model works on data that contains sensitive information. Most privacy-preserving methods lead to undesirable performance degradation. Ensemble learning is an effective way to improve model performance. In this work, we propose a new method for teacher ensembles that uses more informative network outputs under differential private stochastic gradient descent and provide provable privacy guarantees. Out method employs knowledge distillation and hint learning on intermediate represen…
The Private Aggregation of Teacher Ensembles (PATE) is an important private machine learning framework. It combines multiple learning models used as teachers for a student model that learns to predict an output chosen by noisy voting among the teachers. The resulting model satisfies differential privacy and has been shown effective in learning high-quality private models in semisupervised settings or when one wishes to protect the data labels. This paper asks whether this privacy-preserving framework introduces or exacerbates bias and unfairness and shows that PATE can introduce accuracy dispa…
Knowledge distillation in machine learning is the process of transferring knowledge from a large model called the teacher to a smaller model called the student. Knowledge distillation is one of the techniques to compress the large network (teacher) to a smaller network (student) that can be deployed in small devices such as mobile phones. When the network size gap between the teacher and student increases, the performance of the student network decreases. To solve this problem, an intermediate model is employed between the teacher model and the student model known as the teaching assistant mod…
Georgia Institute of Technology（佐治亚理工学院）；Kore University of Enna（恩纳科雷大学）；NTNU（挪威科技大学）
We propose an ensemble learning framework with Poisson sub-sampling to effectively train a collection of teacher models to issue some differential privacy (DP) guarantee for training data. Through boosting under DP, a student model derived from the training data suffers little model degradation from the models trained with no privacy protection. Our proposed solution leverages upon two mechanisms, namely: (i) a privacy budget amplification via Poisson sub-sampling to train a target prediction model that requires less noise to achieve a same level of privacy budget, and (ii) a combination of th…
Modern deep learning has achieved great success in various fields. However, it requires the labeling of huge amounts of data, which is expensive and labor-intensive. Active learning (AL), which identifies the most informative samples to be labeled, is becoming increasingly important to maximize the efficiency of the training process. The existing AL methods mostly use only a single final fixed model for acquiring the samples to be labeled. This strategy may not be good enough in that the structural uncertainty of a model for given training data is not considered to acquire the samples. In this…
Deep learning has contributed greatly to many successes in artificial intelligence in recent years. Today, it is possible to train models that have thousands of layers and hundreds of billions of parameters. Large-scale deep models have achieved great success, but the enormous computational complexity and gigantic storage requirements make it extremely difficult to implement them in real-time applications. On the other hand, the size of the dataset is still a real problem in many domains. Data are often missing, too expensive, or impossible to obtain for other reasons. Ensemble learning is par…
Korea Advanced Institute of Science and Technology（韩国科学技术院）；Naver；AITRICS
Ensembles of deep neural networks have demonstrated superior performance, but their heavy computational cost hinders applying them for resource-limited environments. It motivates distilling knowledge from the ensemble teacher into a smaller student network, and there are two important design choices for this ensemble distillation: 1) how to construct the student network, and 2) what data should be shown during training. In this paper, we propose a weight averaging technique where a student with multiple subnetworks is trained to absorb the functional diversity of ensemble teachers, but then th…
University of Tübingen（蒂宾根大学）；MPI for Informatics（马克斯·普朗克信息学研究所）；MPI for Intelligent Systems（马克斯·普朗克智能系统研究所）；University of Heidelberg（海德堡大学）
Recent advances have indicated the strengths of self-supervised pre-training for improving representation learning on downstream tasks. Existing works often utilize self-supervised pre-trained models by fine-tuning on downstream tasks. However, fine-tuning does not generalize to the case when one needs to build a customized model architecture different from the self-supervised model. In this work, we formulate a new knowledge distillation framework to transfer the knowledge from self-supervised pre-trained models to any other student network by a novel approach named Embedding Graph Alignment.…
Recently, a newly proposed self-supervised framework Bootstrap Your Own Latent (BYOL) seriously challenges the necessity of negative samples in contrastive learning frameworks. BYOL works like a charm despite the fact that it discards the negative samples completely and there is no measure to prevent collapse in its training objective. In this paper, we suggest understanding BYOL from the view of our proposed interpretable self-supervised learning framework, Run Away From your Teacher (RAFT). RAFT optimizes two objectives at the same time: (i) aligning two views of the same data to similar rep…
Seoul National University（首尔大学）；Hanyang University（汉阳大学）
The quantization of deep neural networks (QDNNs) has been actively studied for deployment in edge devices. Recent studies employ the knowledge distillation (KD) method to improve the performance of quantized networks. In this study, we propose stochastic precision ensemble training for QDNNs (SPEQ). SPEQ is a knowledge distillation training scheme; however, the teacher is formed by sharing the model parameters of the student network. We obtain the soft labels of the teacher by changing the bit precision of the activation stochastically at each layer of the forward-pass computation. The student…
Fudan University（复旦大学）；University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
Often the best performing deep neural models are ensembles of multiple base-level networks. Unfortunately, the space required to store these many networks, and the time required to execute them at test-time, prohibits their use in applications where test sets are large (e.g., ImageNet). In this paper, we present a method for compressing large, complex trained ensembles into a single network, where knowledge from a variety of trained deep neural networks (DNNs) is distilled and transferred to a single DNN. In order to distill diverse knowledge from different trained (teacher) models, we propose…
Physics-Informed Teacher-Student Ensemble Learning for Traffic State Estimation with a Varying Speed Limit Scenario
Department of Building, Civil and Environmental Engineering, Concordia University（康科迪亚大学建筑、土木和环境工程系）；Urban Artificial Intelligence Laboratory, University of Florida（佛罗里达大学城市人工智能实验室）；Department of Civil, Environmental and Construction Engineering, University of Central Florida（中央佛罗里达大学土木、环境和建设工程系）
Georgia Institute of Technology（佐治亚理工学院）；University of Enna（恩纳大学）；NTNU（挪威科技大学）
We propose using an adversarial autoencoder (AAE) to replace generative adversarial network (GAN) in the private aggregation of teacher ensembles (PATE), a solution for ensuring differential privacy in speech applications. The AAE architecture allows us to obtain good synthetic speech leveraging upon a discriminative training of latent vectors. Such synthetic speech is used to build a privacy-preserving classifier when non-sensitive data is not sufficiently available in the public domain. This classifier follows the PATE scheme that uses an ensemble of noisy outputs to label the synthetic samp…
Differential privacy (DP) is one data protection avenue to safeguard user information used for training deep models by imposing noisy distortion on privacy data. Such a noise perturbation often results in a severe performance degradation in automatic speech recognition (ASR) in order to meet a privacy budget $\varepsilon$. Private aggregation of teacher ensemble (PATE) utilizes ensemble probabilities to improve ASR accuracy when dealing with the noise effects controlled by small values of $\varepsilon$. We extend PATE learning to work with dynamic patterns, namely speech utterances, and perfor…
Knowledge distillation is an effective approach to learn compact models (students) with the supervision of large and strong models (teachers). As empirically there exists a strong correlation between the performance of teacher and student models, it is commonly believed that a high performing teacher is preferred. Consequently, practitioners tend to use a well trained network or an ensemble of them as the teacher. In this paper, we make an intriguing observation that an intermediate model, i.e., a checkpoint in the middle of the training procedure, often serves as a better teacher compared to …
Syracuse University（雪城大学）；Georgia Institute of Technology（佐治亚理工学院）
A critical concern in data-driven processes is to build models whose outcomes do not discriminate against some demographic groups, including gender, ethnicity, or age. To ensure non-discrimination in learning tasks, knowledge of the group attributes is essential. However, in practice, these attributes may not be available due to legal and ethical requirements. To address this challenge, this paper studies a model that protects the privacy of the individuals' sensitive information while also allowing it to learn non-discriminatory predictors. A key characteristic of the proposed model is to ena…
Siemens AG, Corporate Technology（西门子公司企业技术部）；University of Munich（慕尼黑大学）
In recent years, many explanation methods have been proposed to explain individual classifications of deep neural networks. However, how to leverage the created explanations to improve the learning process has been less explored. As the privileged information, the explanations of a model can be used to guide the learning process of the model itself. In the community, another intensively investigated privileged information used to guide the training of a model is the knowledge from a powerful teacher model. The goal of this work is to leverage the self-explanation to improve the learning proces…
Carnegie Mellon University（卡内基梅隆大学）；University of California San Diego（加利福尼亚大学圣迭戈分校）；Shanghai University of Finance and Economics（上海财经大学）；University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
Generic Image recognition is a fundamental and fairly important visual problem in computer vision. One of the major challenges of this task lies in the fact that single image usually has multiple objects inside while the labels are still one-hot, another one is noisy and sometimes missing labels when annotated by humans. In this paper, we focus on tackling these challenges accompanying with two different image recognition problems: multi-model ensemble and noisy data recognition with a unified framework. As is well-known, usually the best performing deep neural models are ensembles of multiple…
University of Toronto（多伦多大学）；Vector Institute（向量研究所）；University of Oxford（牛津大学）
When learning from sensitive data, care must be taken to ensure that training algorithms address privacy concerns. The canonical Private Aggregation of Teacher Ensembles, or PATE, computes output labels by aggregating the predictions of a (possibly distributed) collection of teacher models via a voting mechanism. The mechanism adds noise to attain a differential privacy guarantee with respect to the teachers' training data. In this work, we observe that this use of noise, which makes PATE predictions stochastic, enables new forms of leakage of sensitive information. For a given input, our adve…
Self-supervised learning on graphs has recently achieved remarkable success in graph representation learning. With hundreds of self-supervised pretext tasks proposed over the past few years, the research community has greatly developed, and the key is no longer to design more powerful but complex pretext tasks, but to make more effective use of those already on hand. This paper studies the problem of how to automatically, adaptively, and dynamically learn instance-level self-supervised learning strategies for each node from a given pool of pretext tasks. In this paper, we propose a novel multi…
Learning from Stochastic Teacher Representations Using Student-Guided Knowledge Distillation
LIVIA, Dept. of Systems Engineering, ETS Montreal（LIVIA，系统工程系，蒙特利尔工程学院）；CentraleSupélec, Universite Paris Saclay（中央理工巴黎高等学院）；Aiim Lab, Queen’s University（人工智能实验室，皇后大学）
Contrastive Learning and Masked Image Modelling have demonstrated exceptional performance on self-supervised representation learning, where Momentum Contrast (i.e., MoCo) and Masked AutoEncoder (i.e., MAE) are the state-of-the-art, respectively. In this work, we propose MOMA to distill from pre-trained MoCo and MAE in a self-supervised manner to collaborate the knowledge from both paradigms. We introduce three different mechanisms of knowledge transfer in the propsoed MOMA framework. : (1) Distill pre-trained MoCo to MAE. (2) Distill pre-trained MAE to MoCo (3) Distill pre-trained MoCo and MAE…
We present Self Meta Pseudo Labels, a novel semi-supervised learning method similar to Meta Pseudo Labels but without the teacher model. We introduce a novel way to use a single model for both generating pseudo labels and classification, allowing us to store only one model in memory instead of two. Our method attains similar performance to the Meta Pseudo Labels method while drastically reducing memory usage.
Georgia Institute of Technology（佐治亚理工学院）；Amazon（亚马逊）；The Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳））
Self-training achieves enormous success in various semi-supervised and weakly-supervised learning tasks. The method can be interpreted as a teacher-student framework, where the teacher generates pseudo-labels, and the student makes predictions. The two models are updated alternatingly. However, such a straightforward alternating update rule leads to training instability. This is because a small change in the teacher may result in a significant change in the student. To address this issue, we propose DRIFT, short for differentiable self-training, that treats teacher-student as a Stackelberg gam…
Knowledge distillation is the procedure of transferring "knowledge" from a large model (the teacher) to a more compact one (the student), often being used in the context of model compression. When both models have the same architecture, this procedure is called self-distillation. Several works have anecdotally shown that a self-distilled student can outperform the teacher on held-out data. In this work, we systematically study self-distillation in a number of settings. We first show that even with a highly accurate teacher, self-distillation allows a student to surpass the teacher in all cases…
Pennsylvania State University（宾夕法尼亚州立大学）；University of California San Diego（加利福尼亚大学圣地亚哥分校）；Google Brain（谷歌大脑）
The rapid adoption of machine learning has increased concerns about the privacy implications of machine learning models trained on sensitive data, such as medical records or other personal information. To address those concerns, one promising approach is Private Aggregation of Teacher Ensembles, or PATE, which transfers to a "student" model the knowledge of an ensemble of "teacher" models, with intuitive privacy provided by training teachers on disjoint data and strong privacy guaranteed by noisy aggregation of teachers' answers. However, PATE has so far been evaluated only on simple classific…
Taiwan AI Labs（台湾人工智能实验室）；National Taiwan University College of Medicine（台湾大学医学院）；Taipei Veterans General Hospital（台北荣民总医院）；Academia Sinica（中央研究院）
Delineation of cancerous regions in gigapixel whole slide images (WSIs) is a crucial diagnostic procedure in digital pathology. This process is time-consuming because of the large search space in the gigapixel WSIs, causing chances of omission and misinterpretation at indistinct tumor lesions. To tackle this, the development of an automated cancerous region segmentation method is imperative. We frame this issue as a modeling problem with partial label WSIs, where some cancerous regions may be misclassified as benign and vice versa, producing patches with noisy labels. To learn from these patch…
