Recent image segmentation models have advanced to segment images into high-quality masks for visual entities, and yet they cannot provide comprehensive semantic understanding for complex queries based on both language and vision. This limitation reduces their effectiveness in applications that require user-friendly interactions driven by vision-language prompts. To bridge this gap, we introduce a novel task of omnimodal referring expression segmentation (ORES). In this task, a model produces a group of masks based on arbitrary prompts specified by text only or text plus reference visual entities. To address this new challenge, we propose a novel framework to "Refer to Any Segmentation Mask Group" (RAS), which augments segmentation models with complex multimodal interactions and comprehension via a mask-centric large multimodal model. For training and benchmarking ORES models, we create datasets MaskGroups-2M and MaskGroups-HQ to include diverse mask groups specified by text and reference entities. Through extensive evaluation, we demonstrate superior performance of RAS on our new ORES task, as well as classic referring expression segmentation (RES) and generalized referring expression segmentation (GRES) tasks. Project page: https://Ref2Any.github.io.

The Segment Anything Model (SAM) is a large-scale foundation model that has revolutionized segmentation methodology. Despite its impressive generalization ability, the segmentation accuracy of SAM on images with intricate structures is often unsatisfactory. Recent works have proposed lightweight fine-tuning using high-quality annotated data to improve accuracy on such images. However, here we provide extensive empirical evidence that this strategy leads to forgetting how to "segment anything": these models lose the original generalization abilities of SAM, in the sense that they perform worse for segmentation tasks not represented in the annotated fine-tuning set. To improve performance without forgetting, we introduce a novel framework that combines high-quality annotated data with a large unlabeled dataset. The framework relies on two methodological innovations. First, we quantify the uncertainty in the SAM pseudo labels associated with the unlabeled data and leverage it to perform uncertainty-aware fine-tuning. Second, we encode the type of segmentation task associated with each training example using a $\textit{task prompt}$ to reduce ambiguity. We evaluated the proposed Segmentation with Uncertainty Model (SUM) on a diverse test set consisting of 14 public benchmarks, where it achieves state-of-the-art results. Notably, our method consistently surpasses SAM by 3-6 points in mean IoU and 4-7 in mean boundary IoU across point-prompt interactive segmentation rounds. Code is available at https://github.com/Kangningthu/SUM

Learning representations for individual instances when only bag-level labels are available is a fundamental challenge in multiple instance learning (MIL). Recent works have shown promising results using contrastive self-supervised learning (CSSL), which learns to push apart representations corresponding to two different randomly-selected instances. Unfortunately, in real-world applications such as medical image classification, there is often class imbalance, so randomly-selected instances mostly belong to the same majority class, which precludes CSSL from learning inter-class differences. To address this issue, we propose a novel framework, Iterative Self-paced Supervised Contrastive Learning for MIL Representations (ItS2CLR), which improves the learned representation by exploiting instance-level pseudo labels derived from the bag-level labels. The framework employs a novel self-paced sampling strategy to ensure the accuracy of pseudo labels. We evaluate ItS2CLR on three medical datasets, showing that it improves the quality of instance-level pseudo labels and representations, and outperforms existing MIL methods in terms of both bag and instance level accuracy.11Code is available at https://github.com/Kangningthu/ItS2CLR

Deep learning in the presence of noisy annotations has been studied extensively in classification, but much less in segmentation tasks. In this work, we study the learning dynamics of deep segmentation networks trained on inaccurately annotated data. We observe a phenomenon that has been previously reported in the context of classification: the networks tend to first fit the clean pixel-level labels during an “early-learning” phase, before eventually memorizing the false annotations. However, in contrast to classification, memorization in segmentation does not arise simultaneously for all semantic categories. Inspired by these findings, we propose a new method for segmentation from noisy annotations with two key elements. First, we detect the beginning of the memorization phase separately for each category during training. This allows us to adaptively correct the noisy annotations in order to exploit early learning. Second, we incorporate a regularization term that enforces consistency across scales to boost robustness against annotation noise. Our method outperforms standard approaches on a medical-imaging segmentation task where noises are synthesized to mimic human annotation errors. It also provides robustness to realistic noisy annotations present in weakly-supervised semantic segmentation, achieving state-of-the-art results on PASCAL VOC 2012.11Code is available at https://github.com/Kangningthu/ADELE

Automatic action identification from video and kinematic data is an important machine learning problem with applications ranging from robotics to smart health. Most existing works focus on identifying coarse actions such as running, climbing, or cutting vegetables, which have relatively long durations and a complex series of motions. This is an important limitation for applications that require identification of more elemental motions at high temporal resolution. For example, in the rehabilitation of arm impairment after stroke, quantifying the training dose (number of repetitions) requires differentiating motions with sub-second durations. Our goal is to bridge this gap. To this end, we introduce a large-scale, multimodal dataset, StrokeRehab, as a new action-recognition benchmark that includes elemental short-duration actions labeled at a high temporal resolution. StrokeRehab consists of a high-quality inertial measurement unit sensor and video data of 51 stroke-impaired patients and 20 healthy subjects performing activities of daily living like feeding, brushing teeth, etc. Because it contains data from both healthy and impaired individuals, StrokeRehab can be used to study the influence of distribution shift in action-recognition tasks. When evaluated on StrokeRehab, current state-of-the-art models for action segmentation produce noisy predictions, which reduces their accuracy in identifying the corresponding sequence of actions. To address this, we propose a novel approach for high-resolution action identification, inspired by speech-recognition techniques, which is based on a sequence-to-sequence model that directly predicts the sequence of actions. This approach outperforms current state-of-the-art methods on StrokeRehab, as well as on the standard benchmark datasets 50Salads, Breakfast, and Jigsaws.

While cross entropy (CE) is the most commonly used loss function to train deep neural networks for classification tasks, many alternative losses have been developed to obtain better empirical performance. Among them, which one is the best to use is still a mystery, because there seem to be multiple factors affecting the answer, such as properties of the dataset, the choice of network architecture, and so on. This paper studies the choice of loss function by examining the last-layer features of deep networks, drawing inspiration from a recent line work showing that the global optimal solution of CE and mean-square-error (MSE) losses exhibits a Neural Collapse phenomenon. That is, for sufficiently large networks trained until convergence, (i) all features of the same class collapse to the corresponding class mean and (ii) the means associated with different classes are in a configuration where their pairwise distances are all equal and maximized. We extend such results and show through global solution and landscape analyses that a broad family of loss functions including commonly used label smoothing (LS) and focal loss (FL) exhibits Neural Collapse. Hence, all relevant losses (i.e., CE, LS, FL, MSE) produce equivalent features on training data. In particular, based on the unconstrained feature model assumption, we provide either the global landscape analysis for LS loss or the local landscape analysis for FL loss and show that the (only!) global minimizers are neural collapse solutions, while all other critical points are strict saddles whose Hessian exhibit negative curvature directions either in the global scope for LS loss or in the local scope for FL loss near the optimal solution. The experiments further show that Neural Collapse features obtained from all relevant losses (i.e., CE, LS, FL, MSE) lead to largely identical performance on test data as well, provided that the network is sufficiently large and trained until convergence.

Existing unsupervised video-to-video translation methods fail to produce translated videos which are frame-wise realistic, semantic information preserving and video-level consistent. In this work, we propose UVIT, a novel unsupervised video-to-video translation model. Our model decomposes the style and the content, uses the specialized encoder-decoder structure and propagates the inter-frame information through bidirectional recurrent neural network (RNN) units. The style-content decomposition mechanism enables us to achieve style consistent video translation results as well as provides us with a good interface for modality flexible translation. In addition, by changing the input frames and style codes incorporated in our translation, we propose a video interpolation loss, which captures temporal information within the sequence to train our building blocks in a self-supervised manner. Our model can produce photo-realistic, spatio-temporal consistent translated videos in a multimodal way. Subjective and objective experimental results validate the superiority of our model over existing methods.

Medical images differ from natural images in significantly higher resolutions and smaller regions of interest. Because of these differences, neural network architectures that work well for natural images might not be applicable to medical image analysis. In this work, we propose a novel neural network model to address these unique properties of medical images. This model first uses a low-capacity, yet memory-efficient, network on the whole image to identify the most informative regions. It then applies another higher-capacity network to collect details from chosen regions. Finally, it employs a fusion module that aggregates global and local information to make a prediction. While existing methods often require lesion segmentation during training, our model is trained with only image-level labels and can generate pixel-level saliency maps indicating possible malignant findings. We apply the model to screening mammography interpretation: predicting the presence or absence of benign and malignant lesions. On the NYU Breast Cancer Screening Dataset, our model outperforms (AUC = 0.93) ResNet-34 and Faster R-CNN in classifying breasts with malignant findings. On the CBIS-DDSM dataset, our model achieves performance (AUC = 0.858) on par with state-of-the-art approaches. Compared to ResNet-34, our model is 4.1x faster for inference while using 78.4% less GPU memory. Furthermore, we demonstrate, in a reader study, that our model surpasses radiologist-level AUC by a margin of 0.11.