Hierarchical Dense Correlation Distillation for Few-Shot Segmentation-Extended Abstract

Few-shot semantic segmentation (FSS) aims to form class-agnostic models segmenting unseen classes with only a handful of annotations. Previous methods limited to the semantic feature and prototype representation suffer from coarse segmentation granularity and train-set overfitting. In this work, we design Hierarchically Decoupled Matching Network (HDMNet) mining pixel-level support correlation based on the transformer architecture. The self-attention modules are used to assist in establishing hierarchical dense features, as a means to accomplish the cascade matching between query and support features. Moreover, we propose a matching module to reduce train-set overfitting and introduce correlation distillation leveraging semantic correspondence from coarse resolution to boost fine-grained segmentation. Our method performs decently in experiments. We achieve 50.0% mIoU on COCO dataset one-shot setting and 56.0% on five-shot segmentation, respectively. The code will be available on the project website. We hope our work can benefit broader industrial applications where novel classes with limited annotations are required to be decently identified.

Temporal Enhanced Training of Multi-view 3D Object Detector via Historical Object Prediction

In this paper, we propose a new paradigm, named Historical Object Prediction (HoP) for multi-view 3D detection to leverage temporal information more effectively. The HoP approach is straightforward: given the current timestamp t, we generate a pseudo Bird's-Eye View (BEV) feature of timestamp t-k from its adjacent frames and utilize this feature to predict the object set at timestamp t-k. Our approach is motivated by the observation that enforcing the detector to capture both the spatial location and temporal motion of objects occurring at historical timestamps can lead to more accurate BEV feature learning. First, we elaborately design short-term and long-term temporal decoders, which can generate the pseudo BEV feature for timestamp t-k without the involvement of its corresponding camera images. Second, an additional object decoder is flexibly attached to predict the object targets using the generated pseudo BEV feature. Note that we only perform HoP during training, thus the proposed method does not introduce extra overheads during inference. As a plug-and-play approach, HoP can be easily incorporated into state-of-the-art BEV detection frameworks, including BEVFormer and BEVDet series. Furthermore, the auxiliary HoP approach is complementary to prevalent temporal modeling methods, leading to significant performance gains. Extensive experiments are conducted to evaluate the effectiveness of the proposed HoP on the nuScenes dataset. We choose the representative methods, including BEVFormer and BEVDet4D-Depth to evaluate our method. Surprisingly, HoP achieves 68.5% NDS and 62.4% mAP with ViT-L on nuScenes test, outperforming all the 3D object detectors on the leaderboard. Codes will be available at https://github.com/Sense-X/HoP.

Hierarchical Dense Correlation Distillation for Few-Shot Segmentation

Few-shot semantic segmentation (FSS) aims to form class-agnostic models segmenting unseen classes with only a handful of annotations. Previous methods limited to the semantic feature and prototype representation suffer from coarse segmentation granularity and train-set overfitting. In this work, we design Hierarchically Decoupled Matching Network (HDMNet) mining pixel-level support correlation based on the transformer architecture. The self-attention modules are used to assist in establishing hierarchical dense features, as a means to accomplish the cascade matching between query and support features. Moreover, we propose a matching module to reduce train-set overfitting and introduce correlation distillation leveraging semantic correspondence from coarse resolution to boost fine-grained segmentation. Our method performs decently in experiments. We achieve $50.0\%$ mIoU on \coco~dataset one-shot setting and $56.0\%$ on five-shot segmentation, respectively.

Two-person Graph Convolutional Network for Skeleton-based Human Interaction Recognition

Graph Convolutional Network (GCN) outperforms previous methods in the skeleton-based human action recognition area, including human-human interaction recognition task. However, when dealing with interaction sequences, current GCN-based methods simply split the two-person skeleton into two discrete sequences and perform graph convolution separately in the manner of single-person action classification. Such operation ignores rich interactive information and hinders effective spatial relationship modeling for semantic pattern learning. To overcome the above shortcoming, we introduce a novel unified two-person graph representing spatial interaction correlations between joints. Also, a properly designed graph labeling strategy is proposed to let our GCN model learn discriminant spatial-temporal interactive features. Experiments show accuracy improvements in both interactions and individual actions when utilizing the proposed two-person graph topology. Finally, we propose a Two-person Graph Convolutional Network (2P-GCN). The proposed 2P-GCN achieves state-of-the-art results on four benchmarks of three interaction datasets, SBU, NTU-RGB+D, and NTU-RGB+D 120.

Elastic Functional Coding of Riemannian Trajectories

Visual observations of dynamic phenomena, such as human actions, are often represented as sequences of smoothly-varying features . In cases where the feature spaces can be structured as Riemannian manifolds, the corresponding representations become trajectories on manifolds. Analysis of these trajectories is challenging due to non-linearity of underlying spaces and high-dimensionality of trajectories. In vision problems, given the nature of physical systems involved, these phenomena are better characterized on a low-dimensional manifold compared to the space of Riemannian trajectories. For instance, if one does not impose physical constraints of the human body, in data involving human action analysis, the resulting representation space will have highly redundant features. Learning an effective, low-dimensional embedding for action representations will have a huge impact in the areas of search and retrieval, visualization, learning, and recognition. The difficulty lies in inherent non-linearity of the domain and temporal variability of actions that can distort any traditional metric between trajectories. To overcome these issues, we use the framework based on transported square-root velocity fields (TSRVF); this framework has several desirable properties, including a rate-invariant metric and vector space representations. We propose to learn an embedding such that each action trajectory is mapped to a single point in a low-dimensional Euclidean space, and the trajectories that differ only in temporal rates map to the same point. We utilize the TSRVF representation, and accompanying statistical summaries of Riemannian trajectories, to extend existing coding methods such as PCA, KSVD and Label Consistent KSVD to Riemannian trajectories or more generally to Riemannian functions.

Video-Based Action Recognition Using Rate-Invariant Analysis of Covariance Trajectories

Statistical classification of actions in videos is mostly performed by extracting relevant features, particularly covariance features, from image frames and studying time series associated with temporal evolutions of these features. A natural mathematical representation of activity videos is in form of parameterized trajectories on the covariance manifold, i.e. the set of symmetric, positive-definite matrices (SPDMs). The variable execution-rates of actions implies variable parameterizations of the resulting trajectories, and complicates their classification. Since action classes are invariant to execution rates, one requires rate-invariant metrics for comparing trajectories. A recent paper represented trajectories using their transported square-root vector fields (TSRVFs), defined by parallel translating scaled-velocity vectors of trajectories to a reference tangent space on the manifold. To avoid arbitrariness of selecting the reference and to reduce distortion introduced during this mapping, we develop a purely intrinsic approach where SPDM trajectories are represented by redefining their TSRVFs at the starting points of the trajectories, and analyzed as elements of a vector bundle on the manifold. Using a natural Riemannain metric on vector bundles of SPDMs, we compute geodesic paths and geodesic distances between trajectories in the quotient space of this vector bundle, with respect to the re-parameterization group. This makes the resulting comparison of trajectories invariant to their re-parameterization. We demonstrate this framework on two applications involving video classification: visual speech recognition or lip-reading and hand-gesture recognition. In both cases we achieve results either comparable to or better than the current literature.