The state-of-the-art unsupervised contrastive visual representation learning methods that have emerged recently (SimCLR, MoCo, SwAV) all make use of data augmentations in order to construct a pretext task of instant discrimination consisting of similar and dissimilar pairs of images. Similar pairs are constructed by randomly extracting patches from the same image and applying several other transformations such as color jittering or blurring, while transformed patches from different image instances in a given batch are regarded as dissimilar pairs. We argue that this approach can result similar pairs that are \textit{semantically} dissimilar. In this work, we address this problem by introducing a \textit{batch curation} scheme that selects batches during the training process that are more inline with the underlying contrastive objective. We provide insights into what constitutes beneficial similar and dissimilar pairs as well as validate \textit{batch curation} on CIFAR10 by integrating it in the SimCLR model.
Carnegie Mellon University（卡内基梅隆大学）；University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
Contrastive learning has been applied successfully to learn vector representations of text. Previous research demonstrated that learning high-quality representations benefits from batch-wise contrastive loss with a large number of negatives. In practice, the technique of in-batch negative is used, where for each example in a batch, other batch examples' positives will be taken as its negatives, avoiding encoding extra negatives. This, however, still conditions each example's loss on all batch examples and requires fitting the entire large batch into GPU memory. This paper introduces a gradient…
the University of Iowa（爱荷华大学）；Google Research（谷歌研究院）；Nanjing University（南京大学）
In this paper, we study contrastive learning from an optimization perspective, aiming to analyze and address a fundamental issue of existing contrastive learning methods that either rely on a large batch size or a large dictionary of feature vectors. We consider a global objective for contrastive learning, which contrasts each positive pair with all negative pairs for an anchor point. From the optimization perspective, we explain why existing methods such as SimCLR require a large batch size in order to achieve a satisfactory result. In order to remove such requirement, we propose a memory-eff…
Contrastive learning is a method of learning visual representations by training Deep Neural Networks (DNNs) to increase the similarity between representations of positive pairs (transformations of the same image) and reduce the similarity between representations of negative pairs (transformations of different images). Here we explore Energy-Based Contrastive Learning (EBCLR) that leverages the power of generative learning by combining contrastive learning with Energy-Based Models (EBMs). EBCLR can be theoretically interpreted as learning the joint distribution of positive pairs, and it shows p…
HSE University（俄罗斯高等经济学院）；Yandex；Qatar Computing Research Institute, HBKU（哈马德·本·哈利法大学卡塔尔计算研究所）
The use of contrastive loss for representation learning has become prominent in computer vision, and it is now getting attention in Natural Language Processing (NLP). Here, we explore the idea of using a batch-softmax contrastive loss when fine-tuning large-scale pre-trained transformer models to learn better task-specific sentence embeddings for pairwise sentence scoring tasks. We introduce and study a number of variations in the calculation of the loss as well as in the overall training procedure; in particular, we find that data shuffling can be quite important. Our experimental results sho…
In this report, we relate the algorithmic design of Barlow Twins' method to the Hilbert-Schmidt Independence Criterion (HSIC), thus establishing it as a contrastive learning approach that is free of negative samples. Through this perspective, we argue that Barlow Twins (and thus the class of negative-sample-free contrastive learning methods) suggests a possibility to bridge the two major families of self-supervised learning philosophies: non-contrastive and contrastive approaches. In particular, Barlow twins exemplified how we could combine the best practices of both worlds: avoiding the need …
NAVER LABS Europe（NAVER LABS 欧洲）；University of Amsterdam（阿姆斯特丹大学）
Contrastive losses have long been a key ingredient of deep metric learning and are now becoming more popular due to the success of self-supervised learning. Recent research has shown the benefit of decomposing such losses into two sub-losses which act in a complementary way when learning the representation network: a positive term and an entropy term. Although the overall loss is thus defined as a combination of two terms, the balance of these two terms is often hidden behind implementation details and is largely ignored and sub-optimal in practice. In this work, we approach the balance of con…
Despite profound successes, contrastive representation learning relies on carefully designed data augmentations using domain specific knowledge. This challenge is magnified in natural language processing where no general rules exist for data augmentation due to the discrete nature of natural language. We tackle this challenge by presenting a Virtual augmentation Supported Contrastive Learning of sentence representations (VaSCL). Originating from the interpretation that data augmentation essentially constructs the neighborhoods of each training instance, we in turn utilize the neighborhood to g…
Duke University（杜克大学）；Microsoft（微软）；Virginia Tech（弗吉尼亚理工大学）；Amazon（亚马逊）；Fudan University（复旦大学）；KAUST（阿卜杜拉国王科技大学）
InfoNCE-based contrastive representation learners, such as SimCLR, have been tremendously successful in recent years. However, these contrastive schemes are notoriously resource demanding, as their effectiveness breaks down with small-batch training (i.e., the log-K curse, whereas K is the batch-size). In this work, we reveal mathematically why contrastive learners fail in the small-batch-size regime, and present a novel simple, non-trivial contrastive objective named FlatNCE, which fixes this issue. Unlike InfoNCE, our FlatNCE no longer explicitly appeals to a discriminative classification go…
Semi-Supervised Contrastive Learning with Orthonormal Prototypes
Department of Electrical Engineering, Statistics, Biostatistics（电气工程系、统计学与生物统计学系）；Wisconsin Institute of Discovery（威斯康星发现研究所）；University of Wisconsin-Madison（威斯康星大学麦迪逊分校）
Bootstrap Your Own Latent (BYOL) is a self-supervised learning approach for image representation. From an augmented view of an image, BYOL trains an online network to predict a target network representation of a different augmented view of the same image. Unlike contrastive methods, BYOL does not explicitly use a repulsion term built from negative pairs in its training objective. Yet, it avoids collapse to a trivial, constant representation. Thus, it has recently been hypothesized that batch normalization (BN) is critical to prevent collapse in BYOL. Indeed, BN flows gradients across batch ele…
Google Research（谷歌研究院）；Boston University（波士顿大学）；Snap Inc.（森宝公司）；MIT（麻省理工学院）
Contrastive learning applied to self-supervised representation learning has seen a resurgence in recent years, leading to state of the art performance in the unsupervised training of deep image models. Modern batch contrastive approaches subsume or significantly outperform traditional contrastive losses such as triplet, max-margin and the N-pairs loss. In this work, we extend the self-supervised batch contrastive approach to the fully-supervised setting, allowing us to effectively leverage label information. Clusters of points belonging to the same class are pulled together in embedding space,…
Hard-Negative Sampling for Contrastive Learning: Optimal Representation Geometry and Neural- vs Dimensional-Collapse
Department of Electrical Engineering, Tufts University（Tufts大学电气工程系）；Department of Engineering, Engineering Technology, East Tennessee State University（东田纳西州立大学工程系）；Department of Electrical and Computer Engineering, Boston University（波士顿大学电气与计算机工程系）
Cadence Design Systems (India) Private Limited（楷登电子科技（印度）私人有限公司）；Indian Institute of Technology, Jodhpur（焦特布尔印度理工学院）
Contrastive learning is a powerful technique to learn representations that are semantically distinctive and geometrically invariant. While most of the earlier approaches have demonstrated its effectiveness on single-modality learning tasks such as image classification, recently there have been a few attempts towards extending this idea to multi-modal data. In this paper, we propose two loss functions based on normalized cross-entropy to perform the task of learning joint visual-semantic embedding using batch contrastive training. In a batch, for a given anchor point from one modality, we consi…
Tsinghua University（清华大学）；Microsoft Research Asia（微软亚洲研究院）
Contrastive learning is widely used for recommendation model learning, where selecting representative and informative negative samples is critical. Existing methods usually focus on centralized data, where abundant and high-quality negative samples are easy to obtain. However, centralized user data storage and exploitation may lead to privacy risks and concerns, while decentralized user data on a single client can be too sparse and biased for accurate contrastive learning. In this paper, we propose a federated contrastive learning method named FedCL for privacy-preserving recommendation, which…
Microsoft Research（微软研究院）；BenevolentAI（BenevolentAI公司）；University of Oxford（牛津大学）；Human Technopole（人类技术园）
Learning meaningful representations of data that can address challenges such as batch effect correction and counterfactual inference is a central problem in many domains including computational biology. Adopting a Conditional VAE framework, we show that marginal independence between the representation and a condition variable plays a key role in both of these challenges. We propose the Contrastive Mixture of Posteriors (CoMP) method that uses a novel misalignment penalty defined in terms of mixtures of the variational posteriors to enforce this independence in latent space. We show that CoMP h…
School of Computing and Information Systems, The University of Melbourne, Australia（墨尔本大学计算机与信息系统学院）；Baskin School of Engineering, University of California, Santa Cruz, USA（加州大学圣克鲁兹分校工程学院）；Institute of Trustworthy Embodied AI, Fudan University, China（复旦大学可信具身人工智能研究所）
Yale University（耶鲁大学）；The University of Texas at Austin（德克萨斯大学奥斯汀分校）
Contrastive learning (CL) aims to learn useful representation without relying on expert annotations in the context of medical image segmentation. Existing approaches mainly contrast a single positive vector (i.e., an augmentation of the same image) against a set of negatives within the entire remainder of the batch by simply mapping all input features into the same constant vector. Despite the impressive empirical performance, those methods have the following shortcomings: (1) it remains a formidable challenge to prevent the collapsing problems to trivial solutions; and (2) we argue that not a…
University of Michigan（密歇根大学）；Amazon Web Services（亚马逊云科技）；Google Cloud AI（谷歌云人工智能）；KAIST（韩国科学技术院）；LG AI Research（LG人工智能研究院）
Contrastive representation learning has shown to be effective to learn representations from unlabeled data. However, much progress has been made in vision domains relying on data augmentations carefully designed using domain knowledge. In this work, we propose i-Mix, a simple yet effective domain-agnostic regularization strategy for improving contrastive representation learning. We cast contrastive learning as training a non-parametric classifier by assigning a unique virtual class to each data in a batch. Then, data instances are mixed in both the input and virtual label spaces, providing mor…
The primary goal of knowledge distillation (KD) is to encapsulate the information of a model learned from a teacher network into a student network, with the latter being more compact than the former. Existing work, e.g., using Kullback-Leibler divergence for distillation, may fail to capture important structural knowledge in the teacher network and often lacks the ability for feature generalization, particularly in situations when teacher and student are built to address different classification tasks. We propose Wasserstein Contrastive Representation Distillation (WCoRD), which leverages both…
Proposal-Contrastive Pretraining for Object Detection from Fewer Data
Université Paris-Saclay（巴黎-萨克雷大学）；CEA（法国原子能委员会）；Université Jean Monnet Saint-Etienne（圣艾蒂安让·莫内大学）；CNRS（法国国家科学研究中心）；Laboratoire Hubert Curien UMR 5516（于贝尔·居里安实验室）；Institut Universitaire de France (IUF)（法国大学研究院）
Graduate School of Artificial Intelligence（人工智能研究生院）；POSTECH（浦项科技大学）
Conditional image generation is the task of generating diverse images using class label information. Although many conditional Generative Adversarial Networks (GAN) have shown realistic results, such methods consider pairwise relations between the embedding of an image and the embedding of the corresponding label (data-to-class relations) as the conditioning losses. In this paper, we propose ContraGAN that considers relations between multiple image embeddings in the same batch (data-to-data relations) as well as the data-to-class relations by using a conditional contrastive loss. The discrimin…
Conditional Neural Processes~(CNPs) bridge neural networks with probabilistic inference to approximate functions of Stochastic Processes under meta-learning settings. Given a batch of non-{\it i.i.d} function instantiations, CNPs are jointly optimized for in-instantiation observation prediction and cross-instantiation meta-representation adaptation within a generative reconstruction pipeline. There can be a challenge in tying together such two targets when the distribution of function observations scales to high-dimensional and noisy spaces. Instead, noise contrastive estimation might be able …
Recent contrastive learning methods achieved state-of-the-art in low label regimes. However, the training requires large batch sizes and heavy augmentations to create multiple views of an image. With non-contrastive methods, the negatives are implicitly incorporated in the loss, allowing different images and modalities as pairs. Although the meta-information (i.e., age, sex) in medical imaging is abundant, the annotations are noisy and prone to class imbalance. In this work, we exploited already existing temporal information (different visits from a patient) in a longitudinal optical coherence…
Mining informative negative instances are of central importance to deep metric learning (DML), however this task is intrinsically limited by mini-batch training, where only a mini-batch of instances is accessible at each iteration. In this paper, we identify a "slow drift" phenomena by observing that the embedding features drift exceptionally slow even as the model parameters are updating throughout the training process. This suggests that the features of instances computed at preceding iterations can be used to considerably approximate their features extracted by the current model. We propose…
Learning Robust Representations for Malicious Content Detection via Contrastive Sampling and Uncertainty Estimation
1 College of Engineering；Computer Science, University of Central Florida, Orlando, USA, 2 Department of Computer Science \& Engineering, Mississippi State University, MS State, USA , 3 Department of Computer Science, University of Nevada, Las Vegas, USA
This paper presents SimCLR: a simple framework for contrastive learning of visual representations. We simplify recently proposed contrastive self-supervised learning algorithms without requiring specialized architectures or a memory bank. In order to understand what enables the contrastive prediction tasks to learn useful representations, we systematically study the major components of our framework. We show that (1) composition of data augmentations plays a critical role in defining effective predictive tasks, (2) introducing a learnable nonlinear transformation between the representation and…
