Fine-grained Text-Video Retrieval with Frozen Image Encoders

State-of-the-art text-video retrieval (TVR) methods typically utilize CLIP and cosine similarity for efficient retrieval. Meanwhile, cross attention methods, which employ a transformer decoder to compute attention between each text query and all frames in a video, offer a more comprehensive interaction between text and videos. However, these methods lack important fine-grained spatial information as they directly compute attention between text and video-level tokens. To address this issue, we propose CrossTVR, a two-stage text-video retrieval architecture. In the first stage, we leverage existing TVR methods with cosine similarity network for efficient text/video candidate selection. In the second stage, we propose a novel decoupled video text cross attention module to capture fine-grained multimodal information in spatial and temporal dimensions. Additionally, we employ the frozen CLIP model strategy in fine-grained retrieval, enabling scalability to larger pre-trained vision models like ViT-G, resulting in improved retrieval performance. Experiments on text video retrieval datasets demonstrate the effectiveness and scalability of our proposed CrossTVR compared to state-of-the-art approaches.

UVOSAM: A Mask-free Paradigm for Unsupervised Video Object Segmentation via Segment Anything Model

Unsupervised video object segmentation has made significant progress in recent years, but the manual annotation of video mask datasets is expensive and limits the diversity of available datasets. The Segment Anything Model (SAM) has introduced a new prompt-driven paradigm for image segmentation, unlocking a range of previously unexplored capabilities. In this paper, we propose a novel paradigm called UVOSAM, which leverages SAM for unsupervised video object segmentation without requiring video mask labels. To address SAM's limitations in instance discovery and identity association, we introduce a video salient object tracking network that automatically generates trajectories for prominent foreground objects. These trajectories then serve as prompts for SAM to produce video masks on a frame-by-frame basis. Our experimental results demonstrate that UVOSAM significantly outperforms current mask-supervised methods. These findings suggest that UVOSAM has the potential to improve unsupervised video object segmentation and reduce the cost of manual annotation.

Monocular Scene Reconstruction with 3D SDF Transformers

Monocular scene reconstruction from posed images is challenging due to the complexity of a large environment. Recent volumetric methods learn to directly predict the TSDF volume and have demonstrated promising results in this task. However, most methods focus on how to extract and fuse the 2D features to a 3D feature volume, but none of them improve the way how the 3D volume is aggregated. In this work, we propose an SDF transformer network, which replaces the role of 3D CNN for better 3D feature aggregation. To reduce the explosive computation complexity of the 3D multi-head attention, we propose a sparse window attention module, where the attention is only calculated between the non-empty voxels within a local window. Then a top-down-bottom-up 3D attention network is built for 3D feature aggregation, where a dilate-attention structure is proposed to prevent geometry degeneration, and two global modules are employed to equip with global receptive fields. The experiments on multiple datasets show that this 3D transformer network generates a more accurate and complete reconstruction, which outperforms previous methods by a large margin. Remarkably, the mesh accuracy is improved by 41.8%, and the mesh completeness is improved by 25.3% on the ScanNet dataset. Project page: https://weihaosky.github.io/sdfformer.

Towards Robust Video Instance Segmentation with Temporal-Aware Transformer

Most existing transformer based video instance segmentation methods extract per frame features independently, hence it is challenging to solve the appearance deformation problem. In this paper, we observe the temporal information is important as well and we propose TAFormer to aggregate spatio-temporal features both in transformer encoder and decoder. Specifically, in transformer encoder, we propose a novel spatio-temporal joint multi-scale deformable attention module which dynamically integrates the spatial and temporal information to obtain enriched spatio-temporal features. In transformer decoder, we introduce a temporal self-attention module to enhance the frame level box queries with the temporal relation. Moreover, TAFormer adopts an instance level contrastive loss to increase the discriminability of instance query embeddings. Therefore the tracking error caused by visually similar instances can be decreased. Experimental results show that TAFormer effectively leverages the spatial and temporal information to obtain context-aware feature representation and outperforms state-of-the-art methods.

Learning Aligned Cross-modal Representations for Referring Image Segmentation

Referring image segmentation aims to segment the image region of interest according to the given language expression, which is a typical multi-modal task. One of the critical challenges of this task is to align semantic representations for different modalities including vision and language. To achieve this, previous methods perform cross-modal interactions to update visual features but ignore the role of integrating fine-grained visual features into linguistic features. We present AlignFormer, an end-to-end framework for referring image segmentation. Our AlignFormer views the linguistic feature as the center embedding and segments the region of interest by pixels grouping based on the center embedding. For achieving the pixel-text alignment, we design a Vision-Language Bidirectional Attention module (VLBA) and resort contrastive learning. Concretely, the VLBA enhances visual features by propagating semantic text representations to each pixel and promotes linguistic features by fusing fine-grained image features. Moreover, we introduce the cross-modal instance contrastive loss to alleviate the influence of pixel samples in ambiguous regions and improve the ability to align multi-modal representations. Extensive experiments demonstrate that our AlignFormer achieves a new state-of-the-art performance on RefCOCO, RefCOCO+, and RefCOCOg by large margins.

RenderNet: Visual Relocalization Using Virtual Viewpoints in Large-Scale Indoor Environments

Visual relocalization has been a widely discussed problem in 3D vision: given a pre-constructed 3D visual map, the 6 DoF (Degrees-of-Freedom) pose of a query image is estimated. Relocalization in large-scale indoor environments enables attractive applications such as augmented reality and robot navigation. However, appearance changes fast in such environments when the camera moves, which is challenging for the relocalization system. To address this problem, we propose a virtual view synthesis-based approach, RenderNet, to enrich the database and refine poses regarding this particular scenario. Instead of rendering real images which requires high-quality 3D models, we opt to directly render the needed global and local features of virtual viewpoints and apply them in the subsequent image retrieval and feature matching operations respectively. The proposed method can largely improve the performance in large-scale indoor environments, e.g., achieving an improvement of 7.1\% and 12.2\% on the Inloc dataset.

RCP: Recurrent Closest Point for Scene Flow Estimation on 3D Point Clouds

3D motion estimation including scene flow and point cloud registration has drawn increasing interest. Inspired by 2D flow estimation, recent methods employ deep neural networks to construct the cost volume for estimating accurate 3D flow. However, these methods are limited by the fact that it is difficult to define a search window on point clouds because of the irregular data structure. In this paper, we avoid this irregularity by a simple yet effective method.We decompose the problem into two interlaced stages, where the 3D flows are optimized point-wisely at the first stage and then globally regularized in a recurrent network at the second stage. Therefore, the recurrent network only receives the regular point-wise information as the input. In the experiments, we evaluate the proposed method on both the 3D scene flow estimation and the point cloud registration task. For 3D scene flow estimation, we make comparisons on the widely used FlyingThings3D and KITTIdatasets. For point cloud registration, we follow previous works and evaluate the data pairs with large pose and partially overlapping from ModelNet40. The results show that our method outperforms the previous method and achieves a new state-of-the-art performance on both 3D scene flow estimation and point cloud registration, which demonstrates the superiority of the proposed zero-order method on irregular point cloud data.

NeW CRFs: Neural Window Fully-connected CRFs for Monocular Depth Estimation

Estimating the accurate depth from a single image is challenging since it is inherently ambiguous and ill-posed. While recent works design increasingly complicated and powerful networks to directly regress the depth map, we take the path of CRFs optimization. Due to the expensive computation, CRFs are usually performed between neighborhoods rather than the whole graph. To leverage the potential of fully-connected CRFs, we split the input into windows and perform the FC-CRFs optimization within each window, which reduces the computation complexity and makes FC-CRFs feasible. To better capture the relationships between nodes in the graph, we exploit the multi-head attention mechanism to compute a multi-head potential function, which is fed to the networks to output an optimized depth map. Then we build a bottom-up-top-down structure, where this neural window FC-CRFs module serves as the decoder, and a vision transformer serves as the encoder. The experiments demonstrate that our method significantly improves the performance across all metrics on both the KITTI and NYUv2 datasets, compared to previous methods. Furthermore, the proposed method can be directly applied to panorama images and outperforms all previous panorama methods on the MatterPort3D dataset. The source code of our method will be made public.

QuadTree Attention for Vision Transformers

Transformers have been successful in many vision tasks, thanks to their capability of capturing long-range dependency. However, their quadratic computational complexity poses a major obstacle for applying them to vision tasks requiring dense predictions, such as object detection, feature matching, stereo, etc. We introduce QuadTree Attention, which reduces the computational complexity from quadratic to linear. Our quadtree transformer builds token pyramids and computes attention in a coarse-to-fine manner. At each level, the top K patches with the highest attention scores are selected, such that at the next level, attention is only evaluated within the relevant regions corresponding to these top K patches. We demonstrate that quadtree attention achieves state-of-the-art performance in various vision tasks, e.g. with 4.0% improvement in feature matching on ScanNet, about 50% flops reduction in stereo matching, 0.4-1.5% improvement in top-1 accuracy on ImageNet classification, 1.2-1.8% improvement on COCO object detection, and 0.7-2.4% improvement on semantic segmentation over previous state-of-the-art transformers. The codes are available at https://github.com/Tangshitao/QuadtreeAttention}{https://github.com/Tangshitao/QuadtreeAttention.

GB-CosFace: Rethinking Softmax-based Face Recognition from the Perspective of Open Set Classification

State-of-the-art face recognition methods typically take the multi-classification pipeline and adopt the softmax-based loss for optimization. Although these methods have achieved great success, the softmax-based loss has its limitation from the perspective of open set classification: the multi-classification objective in the training phase does not strictly match the objective of open set classification testing. In this paper, we derive a new loss named global boundary CosFace (GB-CosFace). Our GB-CosFace introduces an adaptive global boundary to determine whether two face samples belong to the same identity so that the optimization objective is aligned with the testing process from the perspective of open set classification. Meanwhile, since the loss formulation is derived from the softmax-based loss, our GB-CosFace retains the excellent properties of the softmax-based loss, and CosFace is proved to be a special case of the proposed loss. We analyze and explain the proposed GB-CosFace geometrically. Comprehensive experiments on multiple face recognition benchmarks indicate that the proposed GB-CosFace outperforms current state-of-the-art face recognition losses in mainstream face recognition tasks. Compared to CosFace, our GB-CosFace improves 1.58%, 0.57%, and 0.28% at TAR@FAR=1e-6, 1e-5, 1e-4 on IJB-C benchmark.