Parameter-Efficient Fine-Tuning (PEFT) is a fundamental research problem in computer vision, which aims to tune a few parameters for efficient storage and adaptation of pre-trained vision models. Recently, sensitivity-aware parameter efficient fine-tuning method (SPT) addresses this problem by identifying sensitive parameters and then leveraging its sparse characteristic to combine unstructured and structured tuning for PEFT. However, existing methods only focus on the sparse characteristic of sensitive parameters but overlook its distribution characteristic, which results in additional storage burden and limited performance improvement. In this paper, we find that the distribution of sensitive parameters is not chaotic, but concentrates on a small number of rows or columns in each parameter matrix. Inspired by this fact, we propose a Compact Dynamic-Rank Adaptation-based tuning method for Sensitivity-aware Parameter efficient fine-Tuning, called CDRA-SPT. Specifically, we first identify the sensitive parameters that require tuning for each down-stream task. Then, we reorganize the sensitive parameters by following its row and column into a compact sub-parameter matrix. Finally, a dynamic-rank adaptation is designed and applied at sub-parameter matrix level for PEFT. Its advantage is that the dynamic-rank characteristic of sub-parameter matrix can be fully exploited for PEFT. Extensive experiments show that our method achieves superior performance over previous state-of-the-art methods.

In Pedestrian Re-identification (ReID) tasks, extracting robust and discriminative features is a key challenge. Recent studies have mainly concentrated on the extraction of features from single images. However, limited research has been done on considering the relationships between different images, which is crucial in the ReID task. To effectively integrate features from multiple perspectives, we propose an explicitly feature enhancing strategy, Global-local unified cross-view enhancing(GLE) framework for person re-identification, based on Transformer among images sharing the same identity. Specifically, our method includes two aspects: (i) Multi-scale global feature enhancement(MSGE): It displays enhancing global features by interactively updating the features of different images sharing the same identity. (ii) Key local feature enhancement(KLFE): It aims to enhance local feature interaction in key areas by selecting and updating top-k patches based on attention scores derived from thermal maps in each training stage. Experimental results demonstrate that our method can extract more robust and discriminative features, achieving state-of-the-art performance on four pedestrian re-identification datasets.

Symmetric edge polytopes, a.k.a. PV-type adjacency polytopes, associated with undirected graphs have been defined and studied in several seemingly independent areas including number theory, discrete geometry, and dynamical systems. In particular, the authors are motivated by the study of the algebraic Kuramoto equations of unmixed form whose Newton polytopes are the symmetric edge polytopes. The interplay between the geometric structure of symmetric edge polytopes and the topological structure of the underlying graphs has been a recurring theme in recent studies. In particular, “facet/face subgraphs” have emerged as one of the central concepts in describing this symmetry. Continuing along this line of inquiry we provide a complete description of the correspondence between facets/faces of a symmetric edge polytope and maximal bipartite subgraphs of the underlying connected graph.

By using the viewpoint of modern computational algebraic geometry, we explore properties of the optimization landscapes of deep linear neural network models. After providing clarification on the various definitions of “flat” minima, we show that the geometrically flat minima, which are merely artifacts of residual continuous symmetries of the deep linear networks, can be straightforwardly removed by a generalized$L_2$-regularization. Then, we establish upper bounds on the number of isolated stationary points of these networks with the help of algebraic geometry. Combining these upper bounds with a method in numerical algebraic geometry, we findallstationary points for modest depth and matrix size. We demonstrate that, in the presence of the non-zero regularization, deep linear networks can indeed possess local minima which are not global minima. Finally, we show that even though the number of stationary points increases as the number of neurons (regularization parameters) increases (decreases), higher index saddles are surprisingly rare.

Predicting urban congregations can help in monitoring a variety of unusual group events, which is of great importance to public safety and traffic management in smart cities. However, it is very challenging because of complicated spatio-temporal correlations. In this article, we propose a deep neural network-based model, entitled CgNet, for urban congregations prediction. Firstly, we design three types of flows to present dependencies between regions among different timestamps to model the mobility of individuals. Secondly, CgNet utilizes four components, including spatial feature extraction, temporal feature extraction, external factors fusion and congregation feature fusion to collaboratively predict congregations. The combination of these components is capable of not only capturing the spatial and temporal correlations simultaneously, but also learning the essential relationships between three flows and congregations in each region at different stages. Finally, we evaluated the effectiveness of CgNet with extensive experimental study on real taxi trajectory data. The results demonstrate the advantages of our model beyond several baselines.

The pervasiveness of location-acquisition technologies leads to large amounts of spatio-temporal data, which brings us opportunities and challenges to discover interesting group patterns from these individual's trajectories. In this work, firstly, we propose a novel group pattern called congregate group, which captures various congregations by exploiting trajectory streams. Then, we design a discovery framework which contains three main stages including trajectory preprocessing, crowds generation and congregate groups discovery to detect congregations. Meanwhile, an interpolation method is proposed to handle missing points on sparse data. Besides, a set of optimization techniques is applied to reduce computational costs. Finally, our extensive experiments based on real cellular network dataset and real taxicab trajectory dataset demonstrate the effectiveness, efficiency and scalability of our proposed approach.

Belief propagation (BP) is an iterative method to perform approximate inference on arbitrary graphical models. Whether BP converges and if the solution is a unique fixed point depends on both the structure and the parametrization of the model. To understand this dependence it is interesting to find all fixed points. In this work, we formulate a set of polynomial equations, the solutions of which correspond to BP fixed points. To solve such a nonlinear system we present the numerical polynomial-homotopy-continuation (NPHC) method. Experiments on binary Ising models and on error-correcting codes show how our method is capable of obtaining all BP fixed points. On Ising models with fixed parameters we show how the structure influences both the number of fixed points and the convergence properties. We further asses the accuracy of the marginals and weighted combinations thereof. Weighting marginals with their respective partition function increases the accuracy in all experiments. Contrary to the conjecture that uniqueness of BP fixed points implies convergence, we find graphs for which BP fails to converge, even though a unique fixed point exists. Moreover, we show that this fixed point gives a good approximation, and the NPHC method is able to obtain this fixed point.