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, while simultaneously pushing apart clusters of samples from different classes. We analyze two possible versions of the supervised contrastive (SupCon) loss, identifying the best-performing formulation of the loss. On ResNet-200, we achieve top-1 accuracy of 81.4% on the ImageNet dataset, which is 0.8% above the best number reported for this architecture. We show consistent outperformance over cross-entropy on other datasets and two ResNet variants. The loss shows benefits for robustness to natural corruptions, and is more stable to hyperparameter settings such as optimizers and data augmentations. In reduced data settings, it outperforms cross-entropy significantly. Our loss function is simple to implement and reference TensorFlow code is released at https://t.ly/supcon.

While Markov Random Fields (MRFs) are widely used in computer vision, they present a quite challenging inference problem. MRF inference can be accelerated by preprocessing techniques like Dead End Elimination (DEE) [8] or QPBO-based approaches [18, 24, 25] which compute the optimal labeling of a subset of variables. These techniques are guaranteed to never wrongly label a variable but they often leave a large number of variables unlabeled. We address this shortcoming by interpreting pre-processing as a classification problem, which allows us to trade off false positives (i.e., giving a variable an incorrect label) versus false negatives (i.e., failing to label a variable). We describe an efficient discriminative rule that finds optimal solutions for a subset of variables. Our technique provides both per-instance and worst-case guarantees concerning the quality of the solution. Empirical studies were conducted over several benchmark datasets. We obtain a speedup factor of 2 to 12 over expansion moves [4] without preprocessing, and on difficult non-submodular energy functions produce slightly lower energy.

Markov Random Fields (MRFs) are a widely used graphical model, but the inference problem is NP-hard. For first-order MRFs with binary labels, Dead End Elimination (DEE) [7] and QPBO [2, 14] can find the optimal labeling for some variables, the much harder case of larger label sets has been addressed by Kovtun [16, 17] and related methods [12, 23, 24, 25], which impose substantial computational overhead. We describe an efficient algorithm to correctly label a subset of the variables for arbitrary MRFs, with particularly good performance on binary MRFs. We propose a sufficient condition to check if a partial labeling is optimal, which is a generalization of DEE's purely local test. We give a hierarchy of relaxations that provide larger optimal partial labelings at the cost of additional computation. Empirical studies were conducted on several benchmarks, using expansion moves [4] for inference. Our algorithm runs in a few seconds, and improves the speed of MRF inference with expansion moves by a factor of 1.5 to 12.

Graph cuts method such as α-expansion [4] and fusion moves [22] have been successful at solving many optimization problems in computer vision. Higher-order Markov Random Fields (MRF's), which are important for numerous applications, have proven to be very difficult, especially for multilabel MRF's (i.e. more than 2 labels). In this paper we propose a new primal-dual energy minimization method for arbitrary higher-order multilabel MRF's. Primal-dual methods provide guaranteed approximation bounds, and can exploit information in the dual variables to improve their efficiency. Our algorithm generalizes the PD3 [19] technique for first-order MRFs, and relies on a variant of max-flow that can exactly optimize certain higher-order binary MRF's [14]. We provide approximation bounds similar to PD3 [19], and the method is fast in practice. It can optimize non-submodular MRF's, and additionally can in- corporate problem-specific knowledge in the form of fusion proposals. We compare experimentally against the existing approaches that can efficiently handle these difficult energy functions [6, 10, 11]. For higher-order denoising and stereo MRF's, we produce lower energy while running significantly faster.