Stephen Lin

Yutong Chen, Fangyun Wei, Xiao Sun et al. · tsinghua

This paper proposes a simple transfer learning baseline for sign language translation. Existing sign language datasets (e.g. PHOENIX-2014T, CSL-Daily) contain only about 10K-20K pairs of sign videos, gloss annotations and texts, which are an order of magnitude smaller than typical parallel data for training spoken language translation models. Data is thus a bottleneck for training effective sign language translation models. To mitigate this problem, we propose to progressively pretrain the model from general-domain datasets that include a large amount of external supervision to within-domain datasets. Concretely, we pretrain the sign-to-gloss visual network on the general domain of human actions and the within-domain of a sign-to-gloss dataset, and pretrain the gloss-to-text translation network on the general domain of a multilingual corpus and the within-domain of a gloss-to-text corpus. The joint model is fine-tuned with an additional module named the visual-language mapper that connects the two networks. This simple baseline surpasses the previous state-of-the-art results on two sign language translation benchmarks, demonstrating the effectiveness of transfer learning. With its simplicity and strong performance, this approach can serve as a solid baseline for future research. Code and models are available at: https://github.com/FangyunWei/SLRT.

Zhihang Zhong, Xiao Sun, Zhirong Wu et al.

We study the challenging problem of recovering detailed motion from a single motion-blurred image. Existing solutions to this problem estimate a single image sequence without considering the motion ambiguity for each region. Therefore, the results tend to converge to the mean of the multi-modal possibilities. In this paper, we explicitly account for such motion ambiguity, allowing us to generate multiple plausible solutions all in sharp detail. The key idea is to introduce a motion guidance representation, which is a compact quantization of 2D optical flow with only four discrete motion directions. Conditioned on the motion guidance, the blur decomposition is led to a specific, unambiguous solution by using a novel two-stage decomposition network. We propose a unified framework for blur decomposition, which supports various interfaces for generating our motion guidance, including human input, motion information from adjacent video frames, and learning from a video dataset. Extensive experiments on synthesized datasets and real-world data show that the proposed framework is qualitatively and quantitatively superior to previous methods, and also offers the merit of producing physically plausible and diverse solutions. Code is available at https://github.com/zzh-tech/Animation-from-Blur.

Zhihang Zhong, Mingdeng Cao, Xiao Sun et al.

Rolling shutter (RS) distortion can be interpreted as the result of picking a row of pixels from instant global shutter (GS) frames over time during the exposure of the RS camera. This means that the information of each instant GS frame is partially, yet sequentially, embedded into the row-dependent distortion. Inspired by this fact, we address the challenging task of reversing this process, i.e., extracting undistorted GS frames from images suffering from RS distortion. However, since RS distortion is coupled with other factors such as readout settings and the relative velocity of scene elements to the camera, models that only exploit the geometric correlation between temporally adjacent images suffer from poor generality in processing data with different readout settings and dynamic scenes with both camera motion and object motion. In this paper, instead of two consecutive frames, we propose to exploit a pair of images captured by dual RS cameras with reversed RS directions for this highly challenging task. Grounded on the symmetric and complementary nature of dual reversed distortion, we develop a novel end-to-end model, IFED, to generate dual optical flow sequence through iterative learning of the velocity field during the RS time. Extensive experimental results demonstrate that IFED is superior to naive cascade schemes, as well as the state-of-the-art which utilizes adjacent RS images. Most importantly, although it is trained on a synthetic dataset, IFED is shown to be effective at retrieving GS frame sequences from real-world RS distorted images of dynamic scenes. Code is available at https://github.com/zzh-tech/Dual-Reversed-RS.

Huimin Wu, Chenyang Lei, Xiao Sun et al.

Self-supervised representation learning follows a paradigm of withholding some part of the data and tasking the network to predict it from the remaining part. Among many techniques, data augmentation lies at the core for creating the information gap. Towards this end, masking has emerged as a generic and powerful tool where content is withheld along the sequential dimension, e.g., spatial in images, temporal in audio, and syntactic in language. In this paper, we explore the orthogonal channel dimension for generic data augmentation by exploiting precision redundancy. The data for each channel is quantized through a non-uniform quantizer, with the quantized value sampled randomly within randomly sampled quantization bins. From another perspective, quantization is analogous to channel-wise masking, as it removes the information within each bin, but preserves the information across bins. Our approach significantly surpasses existing generic data augmentation methods, while showing on par performance against modality-specific augmentations. We comprehensively evaluate our approach on vision, audio, 3D point clouds, as well as the DABS benchmark which is comprised of various data modalities. The code is available at https: //github.com/microsoft/random_quantize.

Sunghwan Hong, Seokju Cho, Jisu Nam et al. · nvidia, utoronto

This paper presents a novel cost aggregation network, called Volumetric Aggregation with Transformers (VAT), for few-shot segmentation. The use of transformers can benefit correlation map aggregation through self-attention over a global receptive field. However, the tokenization of a correlation map for transformer processing can be detrimental, because the discontinuity at token boundaries reduces the local context available near the token edges and decreases inductive bias. To address this problem, we propose a 4D Convolutional Swin Transformer, where a high-dimensional Swin Transformer is preceded by a series of small-kernel convolutions that impart local context to all pixels and introduce convolutional inductive bias. We additionally boost aggregation performance by applying transformers within a pyramidal structure, where aggregation at a coarser level guides aggregation at a finer level. Noise in the transformer output is then filtered in the subsequent decoder with the help of the query's appearance embedding. With this model, a new state-of-the-art is set for all the standard benchmarks in few-shot segmentation. It is shown that VAT attains state-of-the-art performance for semantic correspondence as well, where cost aggregation also plays a central role.

Atsuhiro Noguchi, Xiao Sun, Stephen Lin et al.

We propose an unsupervised method for 3D geometry-aware representation learning of articulated objects, in which no image-pose pairs or foreground masks are used for training. Though photorealistic images of articulated objects can be rendered with explicit pose control through existing 3D neural representations, these methods require ground truth 3D pose and foreground masks for training, which are expensive to obtain. We obviate this need by learning the representations with GAN training. The generator is trained to produce realistic images of articulated objects from random poses and latent vectors by adversarial training. To avoid a high computational cost for GAN training, we propose an efficient neural representation for articulated objects based on tri-planes and then present a GAN-based framework for its unsupervised training. Experiments demonstrate the efficiency of our method and show that GAN-based training enables the learning of controllable 3D representations without paired supervision.

Jiequan Cui, Yuhui Yuan, Zhisheng Zhong et al.

In this paper, we study the problem of class imbalance in semantic segmentation. We first investigate and identify the main challenges of addressing this issue through pixel rebalance. Then a simple and yet effective region rebalance scheme is derived based on our analysis. In our solution, pixel features belonging to the same class are grouped into region features, and a rebalanced region classifier is applied via an auxiliary region rebalance branch during training. To verify the flexibility and effectiveness of our method, we apply the region rebalance module into various semantic segmentation methods, such as Deeplabv3+, OCRNet, and Swin. Our strategy achieves consistent improvement on the challenging ADE20K and COCO-Stuff benchmark. In particular, with the proposed region rebalance scheme, state-of-the-art BEiT receives +0.7% gain in terms of mIoU on the ADE20K val set.

Chenguo Lin, Xumeng Wen, Wei Cao et al.

Recent research on time-series self-supervised models shows great promise in learning semantic representations. However, it has been limited to small-scale datasets, e.g., thousands of temporal sequences. In this work, we make key technical contributions that are tailored to the numerical properties of time-series data and allow the model to scale to large datasets, e.g., millions of temporal sequences. We adopt the Transformer architecture by first partitioning the input into non-overlapping windows. Each window is then characterized by its normalized shape and two scalar values denoting the mean and standard deviation within each window. To embed scalar values that may possess arbitrary numerical amplitudes in a high-dimensional space, we propose a numerically multi-scaled embedding module enumerating all possible numerical scales for the scalars. The model undergoes pretraining with a simple contrastive objective on a large-scale dataset over a million sequences collected by merging existing public data. We study its transfer performance on a number of univariate and multivariate classification tasks, few shot learning, unsupervised clustering and anomaly detection benchmarks. Our method exhibits remarkable improvement against previous pretraining approaches and establishes the new state of the art, even compared with domain-specific non-learning-based methods. Code is available at: \url{https://github.com/chenguolin/NuTime}.

Sunghwan Hong, Seokju Cho, Seungryong Kim et al. · nvidia, utoronto

We present a novel architecture for dense correspondence. The current state-of-the-art are Transformer-based approaches that focus on either feature descriptors or cost volume aggregation. However, they generally aggregate one or the other but not both, though joint aggregation would boost each other by providing information that one has but other lacks, i.e., structural or semantic information of an image, or pixel-wise matching similarity. In this work, we propose a novel Transformer-based network that interleaves both forms of aggregations in a way that exploits their complementary information. Specifically, we design a self-attention layer that leverages the descriptor to disambiguate the noisy cost volume and that also utilizes the cost volume to aggregate features in a manner that promotes accurate matching. A subsequent cross-attention layer performs further aggregation conditioned on the descriptors of both images and aided by the aggregated outputs of earlier layers. We further boost the performance with hierarchical processing, in which coarser level aggregations guide those at finer levels. We evaluate the effectiveness of the proposed method on dense matching tasks and achieve state-of-the-art performance on all the major benchmarks. Extensive ablation studies are also provided to validate our design choices.

Kun He, Chang Liu, Stephen Lin et al.

In recent years, many data augmentation techniques have been proposed to increase the diversity of input data and reduce the risk of overfitting on deep neural networks. In this work, we propose an easy-to-implement and model-free data augmentation method called Local Magnification (LOMA). Different from other geometric data augmentation methods that perform global transformations on images, LOMA generates additional training data by randomly magnifying a local area of the image. This local magnification results in geometric changes that significantly broaden the range of augmentations while maintaining the recognizability of objects. Moreover, we extend the idea of LOMA and random cropping to the feature space to augment the feature map, which further boosts the classification accuracy considerably. Experiments show that our proposed LOMA, though straightforward, can be combined with standard data augmentation to significantly improve the performance on image classification and object detection. And further combination with our feature augmentation techniques, termed LOMA_IF&FO, can continue to strengthen the model and outperform advanced intensity transformation methods for data augmentation.

Yutong Lin, Ze Liu, Zheng Zhang et al.

Frozen pretrained models have become a viable alternative to the pretraining-then-finetuning paradigm for transfer learning. However, with frozen models there are relatively few parameters available for adapting to downstream tasks, which is problematic in computer vision where tasks vary significantly in input/output format and the type of information that is of value. In this paper, we present a study of frozen pretrained models when applied to diverse and representative computer vision tasks, including object detection, semantic segmentation and video action recognition. From this empirical analysis, our work answers the questions of what pretraining task fits best with this frozen setting, how to make the frozen setting more flexible to various downstream tasks, and the effect of larger model sizes. We additionally examine the upper bound of performance using a giant frozen pretrained model with 3 billion parameters (SwinV2-G) and find that it reaches competitive performance on a varied set of major benchmarks with only one shared frozen base network: 60.0 box mAP and 52.2 mask mAP on COCO object detection test-dev, 57.6 val mIoU on ADE20K semantic segmentation, and 81.7 top-1 accuracy on Kinetics-400 action recognition. With this work, we hope to bring greater attention to this promising path of freezing pretrained image models.

Zhihang Zhong, Mingxi Cheng, Zhirong Wu et al.

Image cropping has progressed tremendously under the data-driven paradigm. However, current approaches do not account for the intentions of the user, which is an issue especially when the composition of the input image is complex. Moreover, labeling of cropping data is costly and hence the amount of data is limited, leading to poor generalization performance of current algorithms in the wild. In this work, we take advantage of vision-language models as a foundation for creating robust and user-intentional cropping algorithms. By adapting a transformer decoder with a pre-trained CLIP-based detection model, OWL-ViT, we develop a method to perform cropping with a text or image query that reflects the user's intention as guidance. In addition, our pipeline design allows the model to learn text-conditioned aesthetic cropping with a small cropping dataset, while inheriting the open-vocabulary ability acquired from millions of text-image pairs. We validate our model through extensive experiments on existing datasets as well as a new cropping test set we compiled that is characterized by content ambiguity.

Zhirong Wu, Zihang Lai, Xiao Sun et al.

The paper presents a scalable approach for learning spatially distributed visual representations over individual tokens and a holistic instance representation simultaneously. We use self-attention blocks to represent spatially distributed tokens, followed by cross-attention blocks to aggregate the holistic image instance. The core of the approach is the use of extremely large token masking (75\%-90\%) as the data augmentation for supervision. Our model, named ExtreMA, follows the plain BYOL approach where the instance representation from the unmasked subset is trained to predict that from the intact input. Instead of encouraging invariance across inputs, the model is required to capture informative variations in an image. The paper makes three contributions: 1) It presents random masking as a strong and computationally efficient data augmentation for siamese representation learning. 2) With multiple sampling per instance, extreme masking greatly speeds up learning and improves performance with more data. 3) ExtreMA obtains stronger linear probing performance than masked modeling methods, and better transfer performance than prior contrastive models.

Yuwei Sun, Hideya Ochiai, Zhirong Wu et al.

Emerging from the pairwise attention in conventional Transformers, there is a growing interest in sparse attention mechanisms that align more closely with localized, contextual learning in the biological brain. Existing studies such as the Coordination method employ iterative cross-attention mechanisms with a bottleneck to enable the sparse association of inputs. However, these methods are parameter inefficient and fail in more complex relational reasoning tasks. To this end, we propose Associative Transformer (AiT) to enhance the association among sparsely attended input tokens, improving parameter efficiency and performance in various vision tasks such as classification and relational reasoning. AiT leverages a learnable explicit memory comprising specialized priors that guide bottleneck attentions to facilitate the extraction of diverse localized tokens. Moreover, AiT employs an associative memory-based token reconstruction using a Hopfield energy function. The extensive empirical experiments demonstrate that AiT requires significantly fewer parameters and attention layers outperforming a broad range of sparse Transformer models. Additionally, AiT outperforms the SOTA sparse Transformer models including the Coordination method on the Sort-of-CLEVR dataset.

Qi Chen, Chao Li, Jia Ning et al.

In convolutional neural networks, the convolutions are conventionally performed using a square kernel with a fixed N $\times$ N receptive field (RF). However, what matters most to the network is the effective receptive field (ERF) that indicates the extent with which input pixels contribute to an output pixel. Inspired by the property that ERFs typically exhibit a Gaussian distribution, we propose a Gaussian Mask convolutional kernel (GMConv) in this work. Specifically, GMConv utilizes the Gaussian function to generate a concentric symmetry mask that is placed over the kernel to refine the RF. Our GMConv can directly replace the standard convolutions in existing CNNs and can be easily trained end-to-end by standard back-propagation. We evaluate our approach through extensive experiments on image classification and object detection tasks. Over several tasks and standard base models, our approach compares favorably against the standard convolution. For instance, using GMConv for AlexNet and ResNet-50, the top-1 accuracy on ImageNet classification is boosted by 0.98% and 0.85%, respectively.

Haoze Li, Jie Zhang, Guoying Zhao et al.

Face Presentation Attack Detection (PAD) demands incremental learning (IL) to combat evolving spoofing tactics and domains. Privacy regulations, however, forbid retaining past data, necessitating rehearsal-free IL (RF-IL). Vision-Language Pre-trained (VLP) models, with their prompt-tunable cross-modal representations, enable efficient adaptation to new spoofing styles and domains. Capitalizing on this strength, we propose \textbf{SVLP-IL}, a VLP-based RF-IL framework that balances stability and plasticity via \textit{Multi-Aspect Prompting} (MAP) and \textit{Selective Elastic Weight Consolidation} (SEWC). MAP isolates domain dependencies, enhances distribution-shift sensitivity, and mitigates forgetting by jointly exploiting universal and domain-specific cues. SEWC selectively preserves critical weights from previous tasks, retaining essential knowledge while allowing flexibility for new adaptations. Comprehensive experiments across multiple PAD benchmarks show that SVLP-IL significantly reduces catastrophic forgetting and enhances performance on unseen domains. SVLP-IL offers a privacy-compliant, practical solution for robust lifelong PAD deployment in RF-IL settings.

Huimin Wu, Kwang-Ting Cheng, Stephen Lin et al.

This paper presents an investigation of vision transformer learning for multi-view geometry tasks, such as optical flow estimation, by fine-tuning video foundation models. Unlike previous methods that involve custom architectural designs and task-specific pretraining, our research finds that general-purpose models pretrained on videos can be readily transferred to multi-view problems with minimal adaptation. The core insight is that general-purpose attention between patches learns temporal and spatial information for geometric reasoning. We demonstrate that appending a linear decoder to the Transformer backbone produces satisfactory results, and iterative refinement can further elevate performance to stateof-the-art levels. This conceptually simple approach achieves top cross-dataset generalization results for optical flow estimation with end-point error (EPE) of 0.69, 1.78, and 3.15 on the Sintel clean, Sintel final, and KITTI datasets, respectively. Our method additionally establishes a new record on the online test benchmark with EPE values of 0.79, 1.88, and F1 value of 3.79. Applications to 3D depth estimation and stereo matching also show strong performance, illustrating the versatility of video-pretrained models in addressing geometric vision tasks.

Mengde Xu, Zheng Zhang, Fangyun Wei et al.

We introduce MixTraining, a new training paradigm for object detection that can improve the performance of existing detectors for free. MixTraining enhances data augmentation by utilizing augmentations of different strengths while excluding the strong augmentations of certain training samples that may be detrimental to training. In addition, it addresses localization noise and missing labels in human annotations by incorporating pseudo boxes that can compensate for these errors. Both of these MixTraining capabilities are made possible through bootstrapping on the detector, which can be used to predict the difficulty of training on a strong augmentation, as well as to generate reliable pseudo boxes thanks to the robustness of neural networks to labeling error. MixTraining is found to bring consistent improvements across various detectors on the COCO dataset. In particular, the performance of Faster R-CNN \cite{ren2015faster} with a ResNet-50 \cite{he2016deep} backbone is improved from 41.7 mAP to 44.0 mAP, and the accuracy of Cascade-RCNN \cite{cai2018cascade} with a Swin-Small \cite{liu2021swin} backbone is raised from 50.9 mAP to 52.8 mAP. The code and models will be made publicly available at \url{https://github.com/MendelXu/MixTraining}.

Ze Liu, Jia Ning, Yue Cao et al.

The vision community is witnessing a modeling shift from CNNs to Transformers, where pure Transformer architectures have attained top accuracy on the major video recognition benchmarks. These video models are all built on Transformer layers that globally connect patches across the spatial and temporal dimensions. In this paper, we instead advocate an inductive bias of locality in video Transformers, which leads to a better speed-accuracy trade-off compared to previous approaches which compute self-attention globally even with spatial-temporal factorization. The locality of the proposed video architecture is realized by adapting the Swin Transformer designed for the image domain, while continuing to leverage the power of pre-trained image models. Our approach achieves state-of-the-art accuracy on a broad range of video recognition benchmarks, including on action recognition (84.9 top-1 accuracy on Kinetics-400 and 86.1 top-1 accuracy on Kinetics-600 with ~20x less pre-training data and ~3x smaller model size) and temporal modeling (69.6 top-1 accuracy on Something-Something v2). The code and models will be made publicly available at https://github.com/SwinTransformer/Video-Swin-Transformer.

Fangyun Wei, Yue Gao, Zhirong Wu et al.

Image-level contrastive representation learning has proven to be highly effective as a generic model for transfer learning. Such generality for transfer learning, however, sacrifices specificity if we are interested in a certain downstream task. We argue that this could be sub-optimal and thus advocate a design principle which encourages alignment between the self-supervised pretext task and the downstream task. In this paper, we follow this principle with a pretraining method specifically designed for the task of object detection. We attain alignment in the following three aspects: 1) object-level representations are introduced via selective search bounding boxes as object proposals; 2) the pretraining network architecture incorporates the same dedicated modules used in the detection pipeline (e.g. FPN); 3) the pretraining is equipped with object detection properties such as object-level translation invariance and scale invariance. Our method, called Selective Object COntrastive learning (SoCo), achieves state-of-the-art results for transfer performance on COCO detection using a Mask R-CNN framework. Code is available at https://github.com/hologerry/SoCo.

Atsuhiro Noguchi, Xiao Sun, Stephen Lin et al.

We present Neural Articulated Radiance Field (NARF), a novel deformable 3D representation for articulated objects learned from images. While recent advances in 3D implicit representation have made it possible to learn models of complex objects, learning pose-controllable representations of articulated objects remains a challenge, as current methods require 3D shape supervision and are unable to render appearance. In formulating an implicit representation of 3D articulated objects, our method considers only the rigid transformation of the most relevant object part in solving for the radiance field at each 3D location. In this way, the proposed method represents pose-dependent changes without significantly increasing the computational complexity. NARF is fully differentiable and can be trained from images with pose annotations. Moreover, through the use of an autoencoder, it can learn appearance variations over multiple instances of an object class. Experiments show that the proposed method is efficient and can generalize well to novel poses. The code is available for research purposes at https://github.com/nogu-atsu/NARF

Ze Liu, Yutong Lin, Yue Cao et al.

This paper presents a new vision Transformer, called Swin Transformer, that capably serves as a general-purpose backbone for computer vision. Challenges in adapting Transformer from language to vision arise from differences between the two domains, such as large variations in the scale of visual entities and the high resolution of pixels in images compared to words in text. To address these differences, we propose a hierarchical Transformer whose representation is computed with \textbf{S}hifted \textbf{win}dows. The shifted windowing scheme brings greater efficiency by limiting self-attention computation to non-overlapping local windows while also allowing for cross-window connection. This hierarchical architecture has the flexibility to model at various scales and has linear computational complexity with respect to image size. These qualities of Swin Transformer make it compatible with a broad range of vision tasks, including image classification (87.3 top-1 accuracy on ImageNet-1K) and dense prediction tasks such as object detection (58.7 box AP and 51.1 mask AP on COCO test-dev) and semantic segmentation (53.5 mIoU on ADE20K val). Its performance surpasses the previous state-of-the-art by a large margin of +2.7 box AP and +2.6 mask AP on COCO, and +3.2 mIoU on ADE20K, demonstrating the potential of Transformer-based models as vision backbones. The hierarchical design and the shifted window approach also prove beneficial for all-MLP architectures. The code and models are publicly available at~\url{https://github.com/microsoft/Swin-Transformer}.

Seokju Lee, Sunghoon Im, Stephen Lin et al.

We present an end-to-end joint training framework that explicitly models 6-DoF motion of multiple dynamic objects, ego-motion and depth in a monocular camera setup without supervision. Our technical contributions are three-fold. First, we highlight the fundamental difference between inverse and forward projection while modeling the individual motion of each rigid object, and propose a geometrically correct projection pipeline using a neural forward projection module. Second, we design a unified instance-aware photometric and geometric consistency loss that holistically imposes self-supervisory signals for every background and object region. Lastly, we introduce a general-purpose auto-annotation scheme using any off-the-shelf instance segmentation and optical flow models to produce video instance segmentation maps that will be utilized as input to our training pipeline. These proposed elements are validated in a detailed ablation study. Through extensive experiments conducted on the KITTI and Cityscapes dataset, our framework is shown to outperform the state-of-the-art depth and motion estimation methods. Our code, dataset, and models are available at https://github.com/SeokjuLee/Insta-DM .

Yue Cao, Jiarui Xu, Stephen Lin et al.

The Non-Local Network (NLNet) presents a pioneering approach for capturing long-range dependencies within an image, via aggregating query-specific global context to each query position. However, through a rigorous empirical analysis, we have found that the global contexts modeled by the non-local network are almost the same for different query positions. In this paper, we take advantage of this finding to create a simplified network based on a query-independent formulation, which maintains the accuracy of NLNet but with significantly less computation. We further replace the one-layer transformation function of the non-local block by a two-layer bottleneck, which further reduces the parameter number considerably. The resulting network element, called the global context (GC) block, effectively models global context in a lightweight manner, allowing it to be applied at multiple layers of a backbone network to form a global context network (GCNet). Experiments show that GCNet generally outperforms NLNet on major benchmarks for various recognition tasks. The code and network configurations are available at https://github.com/xvjiarui/GCNet.

Zhenda Xie, Yutong Lin, Zheng Zhang et al.

Contrastive learning methods for unsupervised visual representation learning have reached remarkable levels of transfer performance. We argue that the power of contrastive learning has yet to be fully unleashed, as current methods are trained only on instance-level pretext tasks, leading to representations that may be sub-optimal for downstream tasks requiring dense pixel predictions. In this paper, we introduce pixel-level pretext tasks for learning dense feature representations. The first task directly applies contrastive learning at the pixel level. We additionally propose a pixel-to-propagation consistency task that produces better results, even surpassing the state-of-the-art approaches by a large margin. Specifically, it achieves 60.2 AP, 41.4 / 40.5 mAP and 77.2 mIoU when transferred to Pascal VOC object detection (C4), COCO object detection (FPN / C4) and Cityscapes semantic segmentation using a ResNet-50 backbone network, which are 2.6 AP, 0.8 / 1.0 mAP and 1.0 mIoU better than the previous best methods built on instance-level contrastive learning. Moreover, the pixel-level pretext tasks are found to be effective for pre-training not only regular backbone networks but also head networks used for dense downstream tasks, and are complementary to instance-level contrastive methods. These results demonstrate the strong potential of defining pretext tasks at the pixel level, and suggest a new path forward in unsupervised visual representation learning. Code is available at \url{https://github.com/zdaxie/PixPro}.

Yihong Chen, Zheng Zhang, Yue Cao et al.

Verification and regression are two general methodologies for prediction in neural networks. Each has its own strengths: verification can be easier to infer accurately, and regression is more efficient and applicable to continuous target variables. Hence, it is often beneficial to carefully combine them to take advantage of their benefits. In this paper, we take this philosophy to improve state-of-the-art object detection, specifically by RepPoints. Though RepPoints provides high performance, we find that its heavy reliance on regression for object localization leaves room for improvement. We introduce verification tasks into the localization prediction of RepPoints, producing RepPoints v2, which provides consistent improvements of about 2.0 mAP over the original RepPoints on the COCO object detection benchmark using different backbones and training methods. RepPoints v2 also achieves 52.1 mAP on COCO \texttt{test-dev} by a single model. Moreover, we show that the proposed approach can more generally elevate other object detection frameworks as well as applications such as instance segmentation. The code is available at https://github.com/Scalsol/RepPointsV2.

Fangyun Wei, Xiao Sun, Hongyang Li et al.

A recent approach for object detection and human pose estimation is to regress bounding boxes or human keypoints from a central point on the object or person. While this center-point regression is simple and efficient, we argue that the image features extracted at a central point contain limited information for predicting distant keypoints or bounding box boundaries, due to object deformation and scale/orientation variation. To facilitate inference, we propose to instead perform regression from a set of points placed at more advantageous positions. This point set is arranged to reflect a good initialization for the given task, such as modes in the training data for pose estimation, which lie closer to the ground truth than the central point and provide more informative features for regression. As the utility of a point set depends on how well its scale, aspect ratio and rotation matches the target, we adopt the anchor box technique of sampling these transformations to generate additional point-set candidates. We apply this proposed framework, called Point-Set Anchors, to object detection, instance segmentation, and human pose estimation. Our results show that this general-purpose approach can achieve performance competitive with state-of-the-art methods for each of these tasks. Code is available at \url{https://github.com/FangyunWei/PointSetAnchor}

Yizhuo Zhang, Zhirong Wu, Houwen Peng et al.

Semi-supervised video object segmentation aims to separate a target object from a video sequence, given the mask in the first frame. Most of current prevailing methods utilize information from additional modules trained in other domains like optical flow and instance segmentation, and as a result they do not compete with other methods on common ground. To address this issue, we propose a simple yet strong transductive method, in which additional modules, datasets, and dedicated architectural designs are not needed. Our method takes a label propagation approach where pixel labels are passed forward based on feature similarity in an embedding space. Different from other propagation methods, ours diffuses temporal information in a holistic manner which take accounts of long-term object appearance. In addition, our method requires few additional computational overhead, and runs at a fast $\sim$37 fps speed. Our single model with a vanilla ResNet50 backbone achieves an overall score of 72.3 on the DAVIS 2017 validation set and 63.1 on the test set. This simple yet high performing and efficient method can serve as a solid baseline that facilitates future research. Code and models are available at \url{https://github.com/microsoft/transductive-vos.pytorch}.

Zhuliang Yao, Yue Cao, Shuxin Zheng et al.

A well-known issue of Batch Normalization is its significantly reduced effectiveness in the case of small mini-batch sizes. When a mini-batch contains few examples, the statistics upon which the normalization is defined cannot be reliably estimated from it during a training iteration. To address this problem, we present Cross-Iteration Batch Normalization (CBN), in which examples from multiple recent iterations are jointly utilized to enhance estimation quality. A challenge of computing statistics over multiple iterations is that the network activations from different iterations are not comparable to each other due to changes in network weights. We thus compensate for the network weight changes via a proposed technique based on Taylor polynomials, so that the statistics can be accurately estimated and batch normalization can be effectively applied. On object detection and image classification with small mini-batch sizes, CBN is found to outperform the original batch normalization and a direct calculation of statistics over previous iterations without the proposed compensation technique. Code is available at https://github.com/Howal/Cross-iterationBatchNorm .

Ze Yang, Yinghao Xu, Han Xue et al.

We present a new object representation, called Dense RepPoints, that utilizes a large set of points to describe an object at multiple levels, including both box level and pixel level. Techniques are proposed to efficiently process these dense points, maintaining near-constant complexity with increasing point numbers. Dense RepPoints is shown to represent and learn object segments well, with the use of a novel distance transform sampling method combined with set-to-set supervision. The distance transform sampling combines the strengths of contour and grid representations, leading to performance that surpasses counterparts based on contours or grids. Code is available at \url{https://github.com/justimyhxu/Dense-RepPoints}.

Seokju Lee, Sunghoon Im, Stephen Lin et al.

We present an end-to-end joint training framework that explicitly models 6-DoF motion of multiple dynamic objects, ego-motion and depth in a monocular camera setup without supervision. Our technical contributions are three-fold. First, we propose a differentiable forward rigid projection module that plays a key role in our instance-wise depth and motion learning. Second, we design an instance-wise photometric and geometric consistency loss that effectively decomposes background and moving object regions. Lastly, we introduce a new auto-annotation scheme to produce video instance segmentation maps that will be utilized as input to our training pipeline. These proposed elements are validated in a detailed ablation study. Through extensive experiments conducted on the KITTI dataset, our framework is shown to outperform the state-of-the-art depth and motion estimation methods. Our code and dataset will be available at https://github.com/SeokjuLee/Insta-DM.

Yue Cao, Jiarui Xu, Stephen Lin et al.

The Non-Local Network (NLNet) presents a pioneering approach for capturing long-range dependencies, via aggregating query-specific global context to each query position. However, through a rigorous empirical analysis, we have found that the global contexts modeled by non-local network are almost the same for different query positions within an image. In this paper, we take advantage of this finding to create a simplified network based on a query-independent formulation, which maintains the accuracy of NLNet but with significantly less computation. We further observe that this simplified design shares similar structure with Squeeze-Excitation Network (SENet). Hence we unify them into a three-step general framework for global context modeling. Within the general framework, we design a better instantiation, called the global context (GC) block, which is lightweight and can effectively model the global context. The lightweight property allows us to apply it for multiple layers in a backbone network to construct a global context network (GCNet), which generally outperforms both simplified NLNet and SENet on major benchmarks for various recognition tasks. The code and configurations are released at https://github.com/xvjiarui/GCNet.

Ze Yang, Shaohui Liu, Han Hu et al.

Modern object detectors rely heavily on rectangular bounding boxes, such as anchors, proposals and the final predictions, to represent objects at various recognition stages. The bounding box is convenient to use but provides only a coarse localization of objects and leads to a correspondingly coarse extraction of object features. In this paper, we present \textbf{RepPoints} (representative points), a new finer representation of objects as a set of sample points useful for both localization and recognition. Given ground truth localization and recognition targets for training, RepPoints learn to automatically arrange themselves in a manner that bounds the spatial extent of an object and indicates semantically significant local areas. They furthermore do not require the use of anchors to sample a space of bounding boxes. We show that an anchor-free object detector based on RepPoints can be as effective as the state-of-the-art anchor-based detection methods, with 46.5 AP and 67.4 $AP_{50}$ on the COCO test-dev detection benchmark, using ResNet-101 model. Code is available at https://github.com/microsoft/RepPoints.

Sunghwan Hong, Seokju Cho, Seungryong Kim et al.

This paper introduces a Transformer-based integrative feature and cost aggregation network designed for dense matching tasks. In the context of dense matching, many works benefit from one of two forms of aggregation: feature aggregation, which pertains to the alignment of similar features, or cost aggregation, a procedure aimed at instilling coherence in the flow estimates across neighboring pixels. In this work, we first show that feature aggregation and cost aggregation exhibit distinct characteristics and reveal the potential for substantial benefits stemming from the judicious use of both aggregation processes. We then introduce a simple yet effective architecture that harnesses self- and cross-attention mechanisms to show that our approach unifies feature aggregation and cost aggregation and effectively harnesses the strengths of both techniques. Within the proposed attention layers, the features and cost volume both complement each other, and the attention layers are interleaved through a coarse-to-fine design to further promote accurate correspondence estimation. Finally at inference, our network produces multi-scale predictions, computes their confidence scores, and selects the most confident flow for final prediction. Our framework is evaluated on standard benchmarks for semantic matching, and also applied to geometric matching, where we show that our approach achieves significant improvements compared to existing methods.

Jie Zhang, Zhifan Wan, Lanqing Hu et al.

Considering the close connection between action recognition and human pose estimation, we design a Collaboratively Self-supervised Video Representation (CSVR) learning framework specific to action recognition by jointly factoring in generative pose prediction and discriminative context matching as pretext tasks. Specifically, our CSVR consists of three branches: a generative pose prediction branch, a discriminative context matching branch, and a video generating branch. Among them, the first one encodes dynamic motion feature by utilizing Conditional-GAN to predict the human poses of future frames, and the second branch extracts static context features by contrasting positive and negative video feature and I-frame feature pairs. The third branch is designed to generate both current and future video frames, for the purpose of collaboratively improving dynamic motion features and static context features. Extensive experiments demonstrate that our method achieves state-of-the-art performance on multiple popular video datasets.

Beibei Lin, Stephen Lin, Robby Tan

Nighttime image dehazing is particularly challenging when dense haze and intense glow severely degrade or entirely obscure background information. Existing methods often struggle due to insufficient background priors and limited generative capability, both of which are highly important under such conditions. In this paper, we introduce BeyondHaze, a generative nighttime dehazing method that not only reduces haze and glow effects but also reconstructs plausible background structures in regions where visual cues are heavily degraded. Our approach is built on two main ideas: obtaining strong background priors by adapting image diffusion models to nighttime dehazing, and enhancing generative ability in haze- and glow-obscured areas through guided training. Task-specific nighttime dehazing knowledge is distilled into an image diffusion model while preserving its capacity to generate clean images. The diffusion model is further trained on tailored image pairs to improve its ability to recover background details that are suppressed by haze effects. Since generative models may introduce hallucinated content, we design our framework to allow user control over the generative level, enabling a balance between visual realism and fidelity. Experiments on real-world nighttime images demonstrate that BeyondHaze substantially improves visibility and scene detail under dense haze.

Kai Qiu, Huishuai Zhang, Zhirong Wu et al.

Training foundation models on extensive datasets and then finetuning them on specific tasks has emerged as the mainstream approach in artificial intelligence. However, the model robustness, which is a critical aspect for safety, is often optimized for each specific task rather than at the pretraining stage. In this paper, we propose a method for pretraining certifiably robust models that can be readily finetuned for adaptation to a particular task. A key challenge is dealing with the compromise between semantic learning and robustness. We address this with a simple yet highly effective strategy based on significantly broadening the pretraining data distribution, which is shown to greatly benefit finetuning for downstream tasks. Through pretraining on a mixture of clean and various noisy images, we find that surprisingly strong certified accuracy can be achieved even when finetuning on only clean images. Furthermore, this strategy requires just a single model to deal with various noise levels, thus substantially reducing computational costs in relation to previous works that employ multiple models. Despite using just one model, our method can still yield results that are on par with, or even superior to, existing multi-model methods.

Shuoxi Zhang, Hanpeng Liu, Stephen Lin et al.

The advent of Vision Transformers (ViTs) marks a substantial paradigm shift in the realm of computer vision. ViTs capture the global information of images through self-attention modules, which perform dot product computations among patchified image tokens. While self-attention modules empower ViTs to capture long-range dependencies, the computational complexity grows quadratically with the number of tokens, which is a major hindrance to the practical application of ViTs. Moreover, the self-attention mechanism in deep ViTs is also susceptible to the attention saturation issue. Accordingly, we argue against the necessity of computing the attention scores in every layer, and we propose the Less-Attention Vision Transformer (LaViT), which computes only a few attention operations at each stage and calculates the subsequent feature alignments in other layers via attention transformations that leverage the previously calculated attention scores. This novel approach can mitigate two primary issues plaguing traditional self-attention modules: the heavy computational burden and attention saturation. Our proposed architecture offers superior efficiency and ease of implementation, merely requiring matrix multiplications that are highly optimized in contemporary deep learning frameworks. Moreover, our architecture demonstrates exceptional performance across various vision tasks including classification, detection and segmentation.

Jie Zhang, Yuhan Li, Yude Wang et al.

Few-shot segmentation (FSS) aims to train a model which can segment the object from novel classes with a few labeled samples. The insufficient generalization ability of models leads to unsatisfactory performance when the models lack enough labeled data from the novel classes. Considering that there are abundant unlabeled data available, it is promising to improve the generalization ability by exploiting these various data. For leveraging unlabeled data, we propose a novel method, named Image to Pseudo-Episode (IPE), to generate pseudo-episodes from unlabeled data. Specifically, our method contains two modules, i.e., the pseudo-label generation module and the episode generation module. The former module generates pseudo-labels from unlabeled images by the spectral clustering algorithm, and the latter module generates pseudo-episodes from pseudo-labeled images by data augmentation methods. Extensive experiments on PASCAL-$5^i$ and COCO-$20^i$ demonstrate that our method achieves the state-of-the-art performance for FSS.

Yinghao Xu, Fangyun Wei, Xiao Sun et al.

Semi-supervised action recognition is a challenging but important task due to the high cost of data annotation. A common approach to this problem is to assign unlabeled data with pseudo-labels, which are then used as additional supervision in training. Typically in recent work, the pseudo-labels are obtained by training a model on the labeled data, and then using confident predictions from the model to teach itself. In this work, we propose a more effective pseudo-labeling scheme, called Cross-Model Pseudo-Labeling (CMPL). Concretely, we introduce a lightweight auxiliary network in addition to the primary backbone, and ask them to predict pseudo-labels for each other. We observe that, due to their different structural biases, these two models tend to learn complementary representations from the same video clips. Each model can thus benefit from its counterpart by utilizing cross-model predictions as supervision. Experiments on different data partition protocols demonstrate the significant improvement of our framework over existing alternatives. For example, CMPL achieves $17.6\%$ and $25.1\%$ Top-1 accuracy on Kinetics-400 and UCF-101 using only the RGB modality and $1\%$ labeled data, outperforming our baseline model, FixMatch, by $9.0\%$ and $10.3\%$, respectively.

Kenneth Li, Xiao Sun, Zhirong Wu et al.

For human action understanding, a popular research direction is to analyze short video clips with unambiguous semantic content, such as jumping and drinking. However, methods for understanding short semantic actions cannot be directly translated to long human dynamics such as dancing, where it becomes challenging even to label the human movements semantically. Meanwhile, the natural language processing (NLP) community has made progress in solving a similar challenge of annotation scarcity by large-scale pre-training, which improves several downstream tasks with one model. In this work, we study how to segment and cluster videos into recurring temporal patterns in a self-supervised way, namely acton discovery, the main roadblock towards video tokenization. We propose a two-stage framework that first obtains a frame-wise representation by contrasting two augmented views of video frames conditioned on their temporal context. The frame-wise representations across a collection of videos are then clustered by K-means. Actons are then automatically extracted by forming a continuous motion sequence from frames within the same cluster. We evaluate the frame-wise representation learning step by Kendall's Tau and the lexicon building step by normalized mutual information and language entropy. We also study three applications of this tokenization: genre classification, action segmentation, and action composition. On the AIST++ and PKU-MMD datasets, actons bring significant performance improvements compared to several baselines.

Runtao Liu, Zhirong Wu, Stella X. Yu et al.

Humans can easily segment moving objects without knowing what they are. That objectness could emerge from continuous visual observations motivates us to model grouping and movement concurrently from unlabeled videos. Our premise is that a video has different views of the same scene related by moving components, and the right region segmentation and region flow would allow mutual view synthesis which can be checked from the data itself without any external supervision. Our model starts with two separate pathways: an appearance pathway that outputs feature-based region segmentation for a single image, and a motion pathway that outputs motion features for a pair of images. It then binds them in a conjoint representation called segment flow that pools flow offsets over each region and provides a gross characterization of moving regions for the entire scene. By training the model to minimize view synthesis errors based on segment flow, our appearance and motion pathways learn region segmentation and flow estimation automatically without building them up from low-level edges or optical flows respectively. Our model demonstrates the surprising emergence of objectness in the appearance pathway, surpassing prior works on zero-shot object segmentation from an image, moving object segmentation from a video with unsupervised test-time adaptation, and semantic image segmentation by supervised fine-tuning. Our work is the first truly end-to-end zero-shot object segmentation from videos. It not only develops generic objectness for segmentation and tracking, but also outperforms prevalent image-based contrastive learning methods without augmentation engineering.

Dong-Jin Kim, Xiao Sun, Jinsoo Choi et al.

A common problem in the task of human-object interaction (HOI) detection is that numerous HOI classes have only a small number of labeled examples, resulting in training sets with a long-tailed distribution. The lack of positive labels can lead to low classification accuracy for these classes. Towards addressing this issue, we observe that there exist natural correlations and anti-correlations among human-object interactions. In this paper, we model the correlations as action co-occurrence matrices and present techniques to learn these priors and leverage them for more effective training, especially on rare classes. The efficacy of our approach is demonstrated experimentally, where the performance of our approach consistently improves over the state-of-the-art methods on both of the two leading HOI detection benchmark datasets, HICO-Det and V-COCO.

Ceyuan Yang, Zhirong Wu, Bolei Zhou et al.

Prior research on self-supervised learning has led to considerable progress on image classification, but often with degraded transfer performance on object detection. The objective of this paper is to advance self-supervised pretrained models specifically for object detection. Based on the inherent difference between classification and detection, we propose a new self-supervised pretext task, called instance localization. Image instances are pasted at various locations and scales onto background images. The pretext task is to predict the instance category given the composited images as well as the foreground bounding boxes. We show that integration of bounding boxes into pretraining promotes better task alignment and architecture alignment for transfer learning. In addition, we propose an augmentation method on the bounding boxes to further enhance the feature alignment. As a result, our model becomes weaker at Imagenet semantic classification but stronger at image patch localization, with an overall stronger pretrained model for object detection. Experimental results demonstrate that our approach yields state-of-the-art transfer learning results for object detection on PASCAL VOC and MSCOCO.

Xin Wei, Guojun Chen, Yue Dong et al.

We present a scheme for fast environment light estimation from the RGBD appearance of individual objects and their local image areas. Conventional inverse rendering is too computationally demanding for real-time applications, and the performance of purely learning-based techniques may be limited by the meager input data available from individual objects. To address these issues, we propose an approach that takes advantage of physical principles from inverse rendering to constrain the solution, while also utilizing neural networks to expedite the more computationally expensive portions of its processing, to increase robustness to noisy input data as well as to improve temporal and spatial stability. This results in a rendering-aware system that estimates the local illumination distribution at an object with high accuracy and in real time. With the estimated lighting, virtual objects can be rendered in AR scenarios with shading that is consistent to the real scene, leading to improved realism.

Ailing Zeng, Xiao Sun, Fuyang Huang et al.

Human poses that are rare or unseen in a training set are challenging for a network to predict. Similar to the long-tailed distribution problem in visual recognition, the small number of examples for such poses limits the ability of networks to model them. Interestingly, local pose distributions suffer less from the long-tail problem, i.e., local joint configurations within a rare pose may appear within other poses in the training set, making them less rare. We propose to take advantage of this fact for better generalization to rare and unseen poses. To be specific, our method splits the body into local regions and processes them in separate network branches, utilizing the property that a joint position depends mainly on the joints within its local body region. Global coherence is maintained by recombining the global context from the rest of the body into each branch as a low-dimensional vector. With the reduced dimensionality of less relevant body areas, the training set distribution within network branches more closely reflects the statistics of local poses instead of global body poses, without sacrificing information important for joint inference. The proposed split-and-recombine approach, called SRNet, can be easily adapted to both single-image and temporal models, and it leads to appreciable improvements in the prediction of rare and unseen poses.

Dong-Jin Kim, Xiao Sun, Jinsoo Choi et al.

A common problem in human-object interaction (HOI) detection task is that numerous HOI classes have only a small number of labeled examples, resulting in training sets with a long-tailed distribution. The lack of positive labels can lead to low classification accuracy for these classes. Towards addressing this issue, we observe that there exist natural correlations and anti-correlations among human-object interactions. In this paper, we model the correlations as action co-occurrence matrices and present techniques to learn these priors and leverage them for more effective training, especially in rare classes. The utility of our approach is demonstrated experimentally, where the performance of our approach exceeds the state-of-the-art methods on both of the two leading HOI detection benchmark datasets, HICO-Det and V-COCO.

Minghao Yin, Zhuliang Yao, Yue Cao et al.

The non-local block is a popular module for strengthening the context modeling ability of a regular convolutional neural network. This paper first studies the non-local block in depth, where we find that its attention computation can be split into two terms, a whitened pairwise term accounting for the relationship between two pixels and a unary term representing the saliency of every pixel. We also observe that the two terms trained alone tend to model different visual clues, e.g. the whitened pairwise term learns within-region relationships while the unary term learns salient boundaries. However, the two terms are tightly coupled in the non-local block, which hinders the learning of each. Based on these findings, we present the disentangled non-local block, where the two terms are decoupled to facilitate learning for both terms. We demonstrate the effectiveness of the decoupled design on various tasks, such as semantic segmentation on Cityscapes, ADE20K and PASCAL Context, object detection on COCO, and action recognition on Kinetics.

Nanxuan Zhao, Zhirong Wu, Rynson W. H. Lau et al.

Contrastive visual pretraining based on the instance discrimination pretext task has made significant progress. Notably, recent work on unsupervised pretraining has shown to surpass the supervised counterpart for finetuning downstream applications such as object detection and segmentation. It comes as a surprise that image annotations would be better left unused for transfer learning. In this work, we investigate the following problems: What makes instance discrimination pretraining good for transfer learning? What knowledge is actually learned and transferred from these models? From this understanding of instance discrimination, how can we better exploit human annotation labels for pretraining? Our findings are threefold. First, what truly matters for the transfer is low-level and mid-level representations, not high-level representations. Second, the intra-category invariance enforced by the traditional supervised model weakens transferability by increasing task misalignment. Finally, supervised pretraining can be strengthened by following an exemplar-based approach without explicit constraints among the instances within the same category.

Nanxuan Zhao, Zhirong Wu, Rynson W. H. Lau et al.

Recent progress in contrastive learning has revolutionized unsupervised representation learning. Concretely, multiple views (augmentations) from the same image are encouraged to map to the similar embeddings, while views from different images are pulled apart. In this paper, through visualizing and diagnosing classification errors, we observe that current contrastive models are ineffective at localizing the foreground object, limiting their ability to extract discriminative high-level features. This is due to the fact that view generation process considers pixels in an image uniformly. To address this problem, we propose a data-driven approach for learning invariance to backgrounds. It first estimates foreground saliency in images and then creates augmentations by copy-and-pasting the foreground onto a variety of backgrounds. The learning still follows the instance discrimination pretext task, so that the representation is trained to disregard background content and focus on the foreground. We study a variety of saliency estimation methods, and find that most methods lead to improvements for contrastive learning. With this approach (DiLo), significant performance is achieved for self-supervised learning on ImageNet classification, and also for object detection on PASCAL VOC and MSCOCO.