arXiv subjects

Li Sun

Publications and source records attributed to Li Sun.

At least 145 records · Page 8Linked to original sources

arXiv2024

Temporal Insight Enhancement: Mitigating Temporal Hallucination in Multimodal Large Language Models

Recent advancements in Multimodal Large Language Models (MLLMs) have significantly enhanced the comprehension of multimedia content, bringing together diverse modalities such as text, images, and videos. However, a critical challenge faced by these models, especially when processing video inputs, is the occurrence of hallucinations - erroneous perceptions or interpretations, particularly at the event level. This study introduces an innovative method to address event-level hallucinations in MLLMs, focusing on specific temporal understanding in video content. Our approach leverages a novel framework that extracts and utilizes event-specific information from both the event query and the provided video to refine MLLMs' response. We propose a unique mechanism that decomposes on-demand event queries into iconic actions. Subsequently, we employ models like CLIP and BLIP2 to predict specific timestamps for event occurrences. Our evaluation, conducted using the Charades-STA dataset, demonstrates a significant reduction in temporal hallucinations and an improvement in the quality of event-related responses. This research not only provides a new perspective in addressing a critical limitation of MLLMs but also contributes a quantitatively measurable method for evaluating MLLMs in the context of temporal-related questions.

cs.CV↗

arXiv2024

Motif-aware Riemannian Graph Neural Network with Generative-Contrastive Learning

Graphs are typical non-Euclidean data of complex structures. In recent years, Riemannian graph representation learning has emerged as an exciting alternative to Euclidean ones. However, Riemannian methods are still in an early stage: most of them present a single curvature (radius) regardless of structural complexity, suffer from numerical instability due to the exponential/logarithmic map, and lack the ability to capture motif regularity. In light of the issues above, we propose the problem of \emph{Motif-aware Riemannian Graph Representation Learning}, seeking a numerically stable encoder to capture motif regularity in a diverse-curvature manifold without labels. To this end, we present a novel Motif-aware Riemannian model with Generative-Contrastive learning (MotifRGC), which conducts a minmax game in Riemannian manifold in a self-supervised manner. First, we propose a new type of Riemannian GCN (D-GCN), in which we construct a diverse-curvature manifold by a product layer with the diversified factor, and replace the exponential/logarithmic map by a stable kernel layer. Second, we introduce a motif-aware Riemannian generative-contrastive learning to capture motif regularity in the constructed manifold and learn motif-aware node representation without external labels. Empirical results show the superiority of MofitRGC.

cs.LG↗

arXiv2024

Contrastive Sequential Interaction Network Learning on Co-Evolving Riemannian Spaces

The sequential interaction network usually find itself in a variety of applications, e.g., recommender system. Herein, inferring future interaction is of fundamental importance, and previous efforts are mainly focused on the dynamics in the classic zero-curvature Euclidean space. Despite the promising results achieved by previous methods, a range of significant issues still largely remains open: On the bipartite nature, is it appropriate to place user and item nodes in one identical space regardless of their inherent difference? On the network dynamics, instead of a fixed curvature space, will the representation spaces evolve when new interactions arrive continuously? On the learning paradigm, can we get rid of the label information costly to acquire? To address the aforementioned issues, we propose a novel Contrastive model for Sequential Interaction Network learning on Co-Evolving RiEmannian spaces, CSINCERE. To the best of our knowledge, we are the first to introduce a couple of co-evolving representation spaces, rather than a single or static space, and propose a co-contrastive learning for the sequential interaction network. In CSINCERE, we formulate a Cross-Space Aggregation for message-passing across representation spaces of different Riemannian geometries, and design a Neural Curvature Estimator based on Ricci curvatures for modeling the space evolvement over time. Thereafter, we present a Reweighed Co-Contrast between the temporal views of the sequential network, so that the couple of Riemannian spaces interact with each other for the interaction prediction without labels. Empirical results on 5 public datasets show the superiority of CSINCERE over the state-of-the-art methods.

cs.LG↗

arXiv2023

Global-correlated 3D-decoupling Transformer for Clothed Avatar Reconstruction

Reconstructing 3D clothed human avatars from single images is a challenging task, especially when encountering complex poses and loose clothing. Current methods exhibit limitations in performance, largely attributable to their dependence on insufficient 2D image features and inconsistent query methods. Owing to this, we present the Global-correlated 3D-decoupling Transformer for clothed Avatar reconstruction (GTA), a novel transformer-based architecture that reconstructs clothed human avatars from monocular images. Our approach leverages transformer architectures by utilizing a Vision Transformer model as an encoder for capturing global-correlated image features. Subsequently, our innovative 3D-decoupling decoder employs cross-attention to decouple tri-plane features, using learnable embeddings as queries for cross-plane generation. To effectively enhance feature fusion with the tri-plane 3D feature and human body prior, we propose a hybrid prior fusion strategy combining spatial and prior-enhanced queries, leveraging the benefits of spatial localization and human body prior knowledge. Comprehensive experiments on CAPE and THuman2.0 datasets illustrate that our method outperforms state-of-the-art approaches in both geometry and texture reconstruction, exhibiting high robustness to challenging poses and loose clothing, and producing higher-resolution textures. Codes will be available at https://github.com/River-Zhang/GTA.

cs.CV↗

arXiv2023

RecursiveDet: End-to-End Region-based Recursive Object Detection

End-to-end region-based object detectors like Sparse R-CNN usually have multiple cascade bounding box decoding stages, which refine the current predictions according to their previous results. Model parameters within each stage are independent, evolving a huge cost. In this paper, we find the general setting of decoding stages is actually redundant. By simply sharing parameters and making a recursive decoder, the detector already obtains a significant improvement. The recursive decoder can be further enhanced by positional encoding (PE) of the proposal box, which makes it aware of the exact locations and sizes of input bounding boxes, thus becoming adaptive to proposals from different stages during the recursion. Moreover, we also design centerness-based PE to distinguish the RoI feature element and dynamic convolution kernels at different positions within the bounding box. To validate the effectiveness of the proposed method, we conduct intensive ablations and build the full model on three recent mainstream region-based detectors. The RecusiveDet is able to achieve obvious performance boosts with even fewer model parameters and slightly increased computation cost. Codes are available at https://github.com/bravezzzzzz/RecursiveDet.

cs.CV↗

arXiv2023

Self-Supervised Continual Graph Learning in Adaptive Riemannian Spaces

Continual graph learning routinely finds its role in a variety of real-world applications where the graph data with different tasks come sequentially. Despite the success of prior works, it still faces great challenges. On the one hand, existing methods work with the zero-curvature Euclidean space, and largely ignore the fact that curvature varies over the coming graph sequence. On the other hand, continual learners in the literature rely on abundant labels, but labeling graph in practice is particularly hard especially for the continuously emerging graphs on-the-fly. To address the aforementioned challenges, we propose to explore a challenging yet practical problem, the self-supervised continual graph learning in adaptive Riemannian spaces. In this paper, we propose a novel self-supervised Riemannian Graph Continual Learner (RieGrace). In RieGrace, we first design an Adaptive Riemannian GCN (AdaRGCN), a unified GCN coupled with a neural curvature adapter, so that Riemannian space is shaped by the learnt curvature adaptive to each graph. Then, we present a Label-free Lorentz Distillation approach, in which we create teacher-student AdaRGCN for the graph sequence. The student successively performs intra-distillation from itself and inter-distillation from the teacher so as to consolidate knowledge without catastrophic forgetting. In particular, we propose a theoretically grounded Generalized Lorentz Projection for the contrastive distillation in Riemannian space. Extensive experiments on the benchmark datasets show the superiority of RieGrace, and additionally, we investigate on how curvature changes over the graph sequence.

cs.LG↗

arXiv2023

From Characters to Words: Hierarchical Pre-trained Language Model for Open-vocabulary Language Understanding

Current state-of-the-art models for natural language understanding require a preprocessing step to convert raw text into discrete tokens. This process known as tokenization relies on a pre-built vocabulary of words or sub-word morphemes. This fixed vocabulary limits the model's robustness to spelling errors and its capacity to adapt to new domains. In this work, we introduce a novel open-vocabulary language model that adopts a hierarchical two-level approach: one at the word level and another at the sequence level. Concretely, we design an intra-word module that uses a shallow Transformer architecture to learn word representations from their characters, and a deep inter-word Transformer module that contextualizes each word representation by attending to the entire word sequence. Our model thus directly operates on character sequences with explicit awareness of word boundaries, but without biased sub-word or word-level vocabulary. Experiments on various downstream tasks show that our method outperforms strong baselines. We also demonstrate that our hierarchical model is robust to textual corruption and domain shift.

cs.CL↗

arXiv2023

SINCERE: Sequential Interaction Networks representation learning on Co-Evolving RiEmannian manifolds

Sequential interaction networks (SIN) have been commonly adopted in many applications such as recommendation systems, search engines and social networks to describe the mutual influence between users and items/products. Efforts on representing SIN are mainly focused on capturing the dynamics of networks in Euclidean space, and recently plenty of work has extended to hyperbolic geometry for implicit hierarchical learning. Previous approaches which learn the embedding trajectories of users and items achieve promising results. However, there are still a range of fundamental issues remaining open. For example, is it appropriate to place user and item nodes in one identical space regardless of their inherent discrepancy? Instead of residing in a single fixed curvature space, how will the representation spaces evolve when new interaction occurs? To explore these issues for sequential interaction networks, we propose SINCERE, a novel method representing Sequential Interaction Networks on Co-Evolving RiEmannian manifolds. SIN- CERE not only takes the user and item embedding trajectories in respective spaces into account, but also emphasizes on the space evolvement that how curvature changes over time. Specifically, we introduce a fresh cross-geometry aggregation which allows us to propagate information across different Riemannian manifolds without breaking conformal invariance, and a curvature estimator which is delicately designed to predict global curvatures effectively according to current local Ricci curvatures. Extensive experiments on several real-world datasets demonstrate the promising performance of SINCERE over the state-of-the-art sequential interaction prediction methods.

cs.LG↗

arXiv2023

Contrastive Graph Clustering in Curvature Spaces

Graph clustering is a longstanding research topic, and has achieved remarkable success with the deep learning methods in recent years. Nevertheless, we observe that several important issues largely remain open. On the one hand, graph clustering from the geometric perspective is appealing but has rarely been touched before, as it lacks a promising space for geometric clustering. On the other hand, contrastive learning boosts the deep graph clustering but usually struggles in either graph augmentation or hard sample mining. To bridge this gap, we rethink the problem of graph clustering from geometric perspective and, to the best of our knowledge, make the first attempt to introduce a heterogeneous curvature space to graph clustering problem. Correspondingly, we present a novel end-to-end contrastive graph clustering model named CONGREGATE, addressing geometric graph clustering with Ricci curvatures. To support geometric clustering, we construct a theoretically grounded Heterogeneous Curvature Space where deep representations are generated via the product of the proposed fully Riemannian graph convolutional nets. Thereafter, we train the graph clusters by an augmentation-free reweighted contrastive approach where we pay more attention to both hard negatives and hard positives in our curvature space. Empirical results on real-world graphs show that our model outperforms the state-of-the-art competitors.

cs.LG↗

arXiv2023

DrasCLR: A Self-supervised Framework of Learning Disease-related and Anatomy-specific Representation for 3D Medical Images

Large-scale volumetric medical images with annotation are rare, costly, and time prohibitive to acquire. Self-supervised learning (SSL) offers a promising pre-training and feature extraction solution for many downstream tasks, as it only uses unlabeled data. Recently, SSL methods based on instance discrimination have gained popularity in the medical imaging domain. However, SSL pre-trained encoders may use many clues in the image to discriminate an instance that are not necessarily disease-related. Moreover, pathological patterns are often subtle and heterogeneous, requiring the ability of the desired method to represent anatomy-specific features that are sensitive to abnormal changes in different body parts. In this work, we present a novel SSL framework, named DrasCLR, for 3D medical imaging to overcome these challenges. We propose two domain-specific contrastive learning strategies: one aims to capture subtle disease patterns inside a local anatomical region, and the other aims to represent severe disease patterns that span larger regions. We formulate the encoder using conditional hyper-parameterized network, in which the parameters are dependant on the anatomical location, to extract anatomically sensitive features. Extensive experiments on large-scale computer tomography (CT) datasets of lung images show that our method improves the performance of many downstream prediction and segmentation tasks. The patient-level representation improves the performance of the patient survival prediction task. We show how our method can detect emphysema subtypes via dense prediction. We demonstrate that fine-tuning the pre-trained model can significantly reduce annotation efforts without sacrificing emphysema detection accuracy. Our ablation study highlights the importance of incorporating anatomical context into the SSL framework.

cs.CV↗

arXiv2023

Approaching Shannon's One-Time Pad: Metrics, Architectures, and Enabling Technologies

The rapid development of advanced computing technologies such as quantum computing imposes new challenges to current wireless security mechanism which is based on cryptographic approaches. To deal with various attacks and realize long-lasting security, we are in urgent need of disruptive security solutions. In this article, novel security transmission paradigms are proposed to approach Shannon's one-time pad perfect secrecy. First, two metrics, termed as Degree-of-Approaching (DoA) and Degree-of-Synchronous-Approaching (DoSA), are developed to characterize the closeness between the achieved security strength and perfect secrecy. These two metrics also serve as a guideline for secure transmission protocol design. After that, we present two paths towards Shannon's one-time pad, i.e., an explicit-encryption based approach and an implicit-encryption based approach. For both of them, we discuss the architecture design, enabling technologies, as well as preliminary performance evaluation results. The techniques presented in this article provide promising security-enhancing solutions for future wireless networks.

cs.CR↗

arXiv2023

MVKT-ECG: Efficient Single-lead ECG Classification on Multi-Label Arrhythmia by Multi-View Knowledge Transferring

The widespread emergence of smart devices for ECG has sparked demand for intelligent single-lead ECG-based diagnostic systems. However, it is challenging to develop a single-lead-based ECG interpretation model for multiple diseases diagnosis due to the lack of some key disease information. In this work, we propose inter-lead Multi-View Knowledge Transferring of ECG (MVKT-ECG) to boost single-lead ECG's ability for multi-label disease diagnosis. This training strategy can transfer superior disease knowledge from multiple different views of ECG (e.g. 12-lead ECG) to single-lead-based ECG interpretation model to mine details in single-lead ECG signals that are easily overlooked by neural networks. MVKT-ECG allows this lead variety as a supervision signal within a teacher-student paradigm, where the teacher observes multi-lead ECG educates a student who observes only single-lead ECG. Since the mutual disease information between the single-lead ECG and muli-lead ECG plays a key role in knowledge transferring, we present a new disease-aware Contrastive Lead-information Transferring(CLT) to improve the mutual disease information between the single-lead ECG and muli-lead ECG. Moreover, We modify traditional Knowledge Distillation to multi-label disease Knowledge Distillation (MKD) to make it applicable for multi-label disease diagnosis. The comprehensive experiments verify that MVKT-ECG has an excellent performance in improving the diagnostic effect of single-lead ECG.

cs.AI↗

arXiv2023

Towards Long-term Autonomy: A Perspective from Robot Learning

In the future, service robots are expected to be able to operate autonomously for long periods of time without human intervention. Many work striving for this goal have been emerging with the development of robotics, both hardware and software. Today we believe that an important underpinning of long-term robot autonomy is the ability of robots to learn on site and on-the-fly, especially when they are deployed in changing environments or need to traverse different environments. In this paper, we examine the problem of long-term autonomy from the perspective of robot learning, especially in an online way, and discuss in tandem its premise "data" and the subsequent "deployment".

cs.RO↗

arXiv2023

Measuring tail risk at high-frequency: An $L_1$-regularized extreme value regression approach with unit-root predictors

We study tail risk dynamics in high-frequency financial markets and their connection with trading activity and market uncertainty. We introduce a dynamic extreme value regression model accommodating both stationary and local unit-root predictors to appropriately capture the time-varying behaviour of the distribution of high-frequency extreme losses. To characterize trading activity and market uncertainty, we consider several volatility and liquidity predictors, and propose a two-step adaptive $L_1$-regularized maximum likelihood estimator to select the most appropriate ones. We establish the oracle property of the proposed estimator for selecting both stationary and local unit-root predictors, and show its good finite sample properties in an extensive simulation study. Studying the high-frequency extreme losses of nine large liquid U.S. stocks using 42 liquidity and volatility predictors, we find the severity of extreme losses to be well predicted by low levels of price impact in period of high volatility of liquidity and volatility.

econ.EM↗

arXiv2023

CLIP-ReID: Exploiting Vision-Language Model for Image Re-Identification without Concrete Text Labels

Pre-trained vision-language models like CLIP have recently shown superior performances on various downstream tasks, including image classification and segmentation. However, in fine-grained image re-identification (ReID), the labels are indexes, lacking concrete text descriptions. Therefore, it remains to be determined how such models could be applied to these tasks. This paper first finds out that simply fine-tuning the visual model initialized by the image encoder in CLIP, has already obtained competitive performances in various ReID tasks. Then we propose a two-stage strategy to facilitate a better visual representation. The key idea is to fully exploit the cross-modal description ability in CLIP through a set of learnable text tokens for each ID and give them to the text encoder to form ambiguous descriptions. In the first training stage, image and text encoders from CLIP keep fixed, and only the text tokens are optimized from scratch by the contrastive loss computed within a batch. In the second stage, the ID-specific text tokens and their encoder become static, providing constraints for fine-tuning the image encoder. With the help of the designed loss in the downstream task, the image encoder is able to represent data as vectors in the feature embedding accurately. The effectiveness of the proposed strategy is validated on several datasets for the person or vehicle ReID tasks. Code is available at https://github.com/Syliz517/CLIP-ReID.

cs.CV↗

arXiv2022

DGFont++: Robust Deformable Generative Networks for Unsupervised Font Generation

Automatic font generation without human experts is a practical and significant problem, especially for some languages that consist of a large number of characters. Existing methods for font generation are often in supervised learning. They require a large number of paired data, which are labor-intensive and expensive to collect. In contrast, common unsupervised image-to-image translation methods are not applicable to font generation, as they often define style as the set of textures and colors. In this work, we propose a robust deformable generative network for unsupervised font generation (abbreviated as DGFont++). We introduce a feature deformation skip connection (FDSC) to learn local patterns and geometric transformations between fonts. The FDSC predicts pairs of displacement maps and employs the predicted maps to apply deformable convolution to the low-level content feature maps. The outputs of FDSC are fed into a mixer to generate final results. Moreover, we introduce contrastive self-supervised learning to learn a robust style representation for fonts by understanding the similarity and dissimilarities of fonts. To distinguish different styles, we train our model with a multi-task discriminator, which ensures that each style can be discriminated independently. In addition to adversarial loss, another two reconstruction losses are adopted to constrain the domain-invariant characteristics between generated images and content images. Taking advantage of FDSC and the adopted loss functions, our model is able to maintain spatial information and generates high-quality character images in an unsupervised manner. Experiments demonstrate that our model is able to generate character images of higher quality than state-of-the-art methods.

cs.CV↗

arXiv2022

A Dual-scale Lead-seperated Transformer With Lead-orthogonal Attention And Meta-information For Ecg Classification

Auxiliary diagnosis of cardiac electrophysiological status can be obtained through the analysis of 12-lead electrocardiograms (ECGs). This work proposes a dual-scale lead-separated transformer with lead-orthogonal attention and meta-information (DLTM-ECG) as a novel approach to address this challenge. ECG segments of each lead are interpreted as independent patches, and together with the reduced dimension signal, they form a dual-scale representation. As a method to reduce interference from segments with low correlation, two group attention mechanisms perform both lead-internal and cross-lead attention. Our method allows for the addition of previously discarded meta-information, further improving the utilization of clinical information. Experimental results show that our DLTM-ECG yields significantly better classification scores than other transformer-based models,matching or performing better than state-of-the-art (SOTA) deep learning methods on two benchmark datasets. Our work has the potential for similar multichannel bioelectrical signal processing and physiological multimodal tasks.

cs.CV↗

arXiv2022

The Cassiopeia Filament: A Blown Spur of the Local Arm

We present wide-field and high-sensitivity CO(1-0) molecular line observations toward the Cassiopeia region, using the 13.7m millimeter telescope of the Purple Mountain Observatory (PMO). The CO observations reveal a large-scale highly filamentary molecular cloud within the Galactic region of 132\fdg0\,$\geq$\,$l$\,$\geq$\,122\fdg0 and -1\fdg0\,$\leq$\,$b$\,$\leq$\,3\fdg0 and the velocity range from approximately +1 to +4 km/s. The measured length of the large-scale filament, referred to as the Cassiopeia Filament, is about 390 pc. The observed properties of the Cassiopeia Filament, such as length, column density, and velocity gradient, are consistent with those synthetic large-scale filaments in the inter-arm regions. Based on its observed properties and location on the Galactic plane, we suggest that the Cassiopeia Filament is a spur of the Local arm, which is formed due to the galactic shear. The western end of the Cassiopeia Filament shows a giant arc-like molecular gas shell, which is extending in the velocity range from roughly -1 to +7 km/s. Finger-like structures, with systematic velocity gradients, are detected in the shell. The CO kinematics suggest that the large shell is expanding at a velocity of ~6.5 km/s. Both the shell and finger-like structures outline a giant bubble with a radius of ~16 pc, which is likely produced by stellar wind from the progenitor star of a supernova remnant. The observed spectral linewidths suggest that the whole Cassiopeia Filament was quiescent initially until its west part was blown by stellar wind and became supersonically turbulent.

astro-ph.GA↗