arXiv subjects

Tong Zhang

Publications and source records attributed to Tong Zhang.

At least 253 records · Page 14Linked to original sources

arXiv2025

MMHCL: Multi-Modal Hypergraph Contrastive Learning for Recommendation

The burgeoning presence of multimodal content-sharing platforms propels the development of personalized recommender systems. Previous works usually suffer from data sparsity and cold-start problems, and may fail to adequately explore semantic user-product associations from multimodal data. To address these issues, we propose a novel Multi-Modal Hypergraph Contrastive Learning (MMHCL) framework for user recommendation. For a comprehensive information exploration from user-product relations, we construct two hypergraphs, i.e. a user-to-user (u2u) hypergraph and an item-to-item (i2i) hypergraph, to mine shared preferences among users and intricate multimodal semantic resemblance among items, respectively. This process yields denser second-order semantics that are fused with first-order user-item interaction as complementary to alleviate the data sparsity issue. Then, we design a contrastive feature enhancement paradigm by applying synergistic contrastive learning. By maximizing/minimizing the mutual information between second-order (e.g. shared preference pattern for users) and first-order (information of selected items for users) embeddings of the same/different users and items, the feature distinguishability can be effectively enhanced. Compared with using sparse primary user-item interaction only, our MMHCL obtains denser second-order hypergraphs and excavates more abundant shared attributes to explore the user-product associations, which to a certain extent alleviates the problems of data sparsity and cold-start. Extensive experiments have comprehensively demonstrated the effectiveness of our method. Our code is publicly available at: https://github.com/Xu107/MMHCL.

cs.IR↗

arXiv2025

Reimagining Memory Access for LLM Inference: Compression-Aware Memory Controller Design

The efficiency of Large Language Model~(LLM) inference is often constrained by substantial memory bandwidth and capacity demands. Existing techniques, such as pruning, quantization, and mixture of experts/depth, reduce memory capacity and/or bandwidth consumption at the cost of slight degradation in inference quality. This paper introduces a design solution that further alleviates memory bottlenecks by enhancing the on-chip memory controller in AI accelerators to achieve two main objectives: (1) significantly reducing memory capacity and bandwidth usage through lossless block compression~(e.g., LZ4 and ZSTD) of model weights and key-value (KV) cache without compromising inference quality, and (2) enabling memory bandwidth and energy consumption to scale proportionally with context-dependent dynamic quantization. These goals are accomplished by equipping the on-chip memory controller with mechanisms to improve fine-grained bit-level accessibility and compressibility of weights and KV cache through LLM-aware configuration of in-memory placement and representation. Experimental results on publicly available LLMs demonstrate the effectiveness of this approach, showing memory footprint reductions of 25.2\% for model weights and 46.9\% for KV cache. In addition, our hardware prototype at 4\,GHz and 32 lanes (7\,nm) achieves 8\,TB/s throughput with a modest area overhead (under 3.8\,mm$^2$), which underscores the viability of LLM-aware memory control as a key to efficient large-scale inference.

cs.AR↗

arXiv2025

Fox-1: Open Small Language Model for Cloud and Edge

We present Fox-1, a series of small language models (SLMs) consisting of Fox-1-1.6B and Fox-1-1.6B-Instruct-v0.1. These models are pre-trained on 3 trillion tokens of web-scraped document data and fine-tuned with 5 billion tokens of instruction-following and multi-turn conversation data. Aiming to improve the pre-training efficiency, Fox-1-1.6B model introduces a novel 3-stage data curriculum across all the training data with 2K-8K sequence length. In architecture design, Fox-1 features a deeper layer structure, an expanded vocabulary, and utilizes Grouped Query Attention (GQA), offering a performant and efficient architecture compared to other SLMs. Fox-1 achieves better or on-par performance in various benchmarks compared to StableLM-2-1.6B, Gemma-2B, Qwen1.5-1.8B, and OpenELM1.1B, with competitive inference speed and throughput. The model weights have been released under the Apache 2.0 license, where we aim to promote the democratization of LLMs and make them fully accessible to the whole open-source community.

cs.CL↗

arXiv2025

AdaptRec: A Self-Adaptive Framework for Sequential Recommendations with Large Language Models

The recent advancements in Large Language Models (LLMs) have generated considerable interest in their utilization for sequential recommendation tasks. While collaborative signals from similar users are central to recommendation modeling, effectively transforming these signals into a format that LLMs can understand and utilize remains challenging. The critical challenges include selecting relevant demonstrations from large-scale user interactions and ensuring their alignment with LLMs' reasoning process. To address these challenges, we introduce AdaptRec, a self-adaptive fram-ework that leverages LLMs for sequential recommendations by incorporating explicit collaborative signals. AdaptRec employs a two-phase user selection mechanism -- User Similarity Retrieval and Self-Adaptive User Selection -- to efficiently identify relevant user sequences in large-scale datasets from multi-metric evaluation. We also develop a User-Based Similarity Retrieval Prompt, enabling the model to actively select similar users and continuously refine its selection criteria during training. Using the collaborative signals from similar users, we construct a User-Contextualized Recommendation Prompt that translates their behavior sequences into natural language, explicitly integrating this information into the recommendation process. Experiments demonstrate AdaptRec's superior performance, with significant improvements in HitRatio@1 scores of 7.13\%, 18.16\%, and 10.41\% across real-world datasets with full fine-tuning, and even higher gains of 23.00\%, 15.97\%, and 17.98\% in few-shot scenarios.

cs.IR↗

arXiv2025

Refining CLIP's Spatial Awareness: A Visual-Centric Perspective

Contrastive Language-Image Pre-training (CLIP) excels in global alignment with language but exhibits limited sensitivity to spatial information, leading to strong performance in zero-shot classification tasks but underperformance in tasks requiring precise spatial understanding. Recent approaches have introduced Region-Language Alignment (RLA) to enhance CLIP's performance in dense multimodal tasks by aligning regional visual representations with corresponding text inputs. However, we find that CLIP ViTs fine-tuned with RLA suffer from notable loss in spatial awareness, which is crucial for dense prediction tasks. To address this, we propose the Spatial Correlation Distillation (SCD) framework, which preserves CLIP's inherent spatial structure and mitigates the above degradation. To further enhance spatial correlations, we introduce a lightweight Refiner that extracts refined correlations directly from CLIP before feeding them into SCD, based on an intriguing finding that CLIP naturally captures high-quality dense features. Together, these components form a robust distillation framework that enables CLIP ViTs to integrate both visual-language and visual-centric improvements, achieving state-of-the-art results across various open-vocabulary dense prediction benchmarks.

cs.CV↗

arXiv2025

VGRP-Bench: Visual Grid Reasoning Puzzle Benchmark for Large Vision-Language Models

Large Vision-Language Models (LVLMs) struggle with puzzles, which require precise perception, rule comprehension, and logical reasoning. Assessing and enhancing their performance in this domain is crucial, as it reflects their ability to engage in structured reasoning - an essential skill for real-world problem-solving. However, existing benchmarks primarily evaluate pre-trained models without additional training or fine-tuning, often lack a dedicated focus on reasoning, and fail to establish a systematic evaluation framework. To address these limitations, we introduce VGRP-Bench, a Visual Grid Reasoning Puzzle Benchmark featuring 20 diverse puzzles. VGRP-Bench spans multiple difficulty levels, and includes extensive experiments not only on existing chat LVLMs (e.g., GPT-4o), but also on reasoning LVLMs (e.g., Gemini-Thinking). Our results reveal that even the state-of-the-art LVLMs struggle with these puzzles, highlighting fundamental limitations in their puzzle-solving capabilities. Most importantly, through systematic experiments, we identify and analyze key factors influencing LVLMs' puzzle-solving performance, including the number of clues, grid size, and rule complexity. Furthermore, we explore two Supervised Fine-Tuning (SFT) strategies that can be used in post-training: SFT on solutions (S-SFT) and SFT on synthetic reasoning processes (R-SFT). While both methods significantly improve performance on trained puzzles, they exhibit limited generalization to unseen ones. We will release VGRP-Bench to facilitate further research on LVLMs for complex, real-world problem-solving. Project page: https://yufan-ren.com/subpage/VGRP-Bench/.

cs.CV↗

arXiv2025

Adaptive Multi-step Refinement Network for Robust Point Cloud Registration

Point Cloud Registration (PCR) estimates the relative rigid transformation between two point clouds of the same scene. Despite significant progress with learning-based approaches, existing methods still face challenges when the overlapping region between the two point clouds is small. In this paper, we propose an adaptive multi-step refinement network that refines the registration quality at each step by leveraging the information from the preceding step. To achieve this, we introduce a training procedure and a refinement network. Firstly, to adapt the network to the current step, we utilize a generalized one-way attention mechanism, which prioritizes the last step's estimated overlapping region, and we condition the network on step indices. Secondly, instead of training the network to map either random transformations or a fixed pre-trained model's estimations to the ground truth, we train it on transformations with varying registration qualities, ranging from accurate to inaccurate, thereby enhancing the network's adaptiveness and robustness. Despite its conceptual simplicity, our method achieves state-of-the-art performance on both the 3DMatch/3DLoMatch and KITTI benchmarks. Notably, on 3DLoMatch, our method reaches 80.4% recall rate, with an absolute improvement of 1.2%.

cs.CV↗

arXiv2025

Performance Characterizations and Usage Guidelines of Samsung CXL Memory Module Hybrid Prototype

The growing prevalence of data-intensive workloads, such as artificial intelligence (AI), machine learning (ML), high-performance computing (HPC), in-memory databases, and real-time analytics, has exposed limitations in conventional memory technologies like DRAM. While DRAM offers low latency and high throughput, it is constrained by high costs, scalability challenges, and volatility, making it less viable for capacity-bound and persistent applications in modern datacenters. Recently, Compute Express Link (CXL) has emerged as a promising alternative, enabling high-speed, cacheline-granular communication between CPUs and external devices. By leveraging CXL technology, NAND flash can now be used as memory expansion, offering three-fold benefits: byte-addressability, scalable capacity, and persistence at a low cost. Samsung's CXL Memory Module Hybrid (CMM-H) is the first product to deliver these benefits through a hardware-only solution, i.e., it does not incur any OS and IO overheads like conventional block devices. In particular, CMM-H integrates a DRAM cache with NAND flash in a single device to deliver near-DRAM latency. This paper presents the first publicly available study for comprehensive characterizations of an FPGA-based CMM-H prototype. Through this study, we address users' concerns about whether a wide variety of applications can successfully run on a memory device backed by NAND flash medium. Additionally, based on these characterizations, we provide key insights into how to best take advantage of the CMM-H device.

cs.AR↗

arXiv2025

Generating Multimodal Driving Scenes via Next-Scene Prediction

Generative models in Autonomous Driving (AD) enable diverse scene creation, yet existing methods fall short by only capturing a limited range of modalities, restricting the capability of generating controllable scenes for comprehensive evaluation of AD systems. In this paper, we introduce a multimodal generation framework that incorporates four major data modalities, including a novel addition of map modality. With tokenized modalities, our scene sequence generation framework autoregressively predicts each scene while managing computational demands through a two-stage approach. The Temporal AutoRegressive (TAR) component captures inter-frame dynamics for each modality while the Ordered AutoRegressive (OAR) component aligns modalities within each scene by sequentially predicting tokens in a fixed order. To maintain coherence between map and ego-action modalities, we introduce the Action-aware Map Alignment (AMA) module, which applies a transformation based on the ego-action to maintain coherence between these modalities. Our framework effectively generates complex, realistic driving scenes over extended sequences, ensuring multimodal consistency and offering fine-grained control over scene elements. Project page: https://yanhaowu.github.io/UMGen/

cs.CV↗

arXiv2025

Automated UI Interface Generation via Diffusion Models: Enhancing Personalization and Efficiency

This study proposes a UI interface generation method based on a diffusion model, aiming to achieve high-quality, diversified, and personalized interface design through generative artificial intelligence technology. The diffusion model is based on its step-by-step denoising generation process. By combining the conditional generation mechanism, design optimization module, and user feedback mechanism, the model can generate a UI interface that meets the requirements based on multimodal inputs such as text descriptions and sketches provided by users. In the study, a complete experimental evaluation framework was designed, and mainstream generation models (such as GAN, VAE, DALL E, etc.) were selected for comparative experiments. The generation results were quantitatively analyzed from indicators such as PSNR, SSIM, and FID. The results show that the model proposed in this study is superior to other models in terms of generation quality and user satisfaction, especially in terms of logical clarity of information transmission and visual aesthetics. The ablation experiment further verifies the key role of conditional generation and design optimization modules in improving interface quality. This study provides a new technical path for UI design automation and lays the foundation for the intelligent and personalized development of human-computer interaction interfaces. In the future, the application potential of the model in virtual reality, game design, and other fields will be further explored.

cs.HC↗

arXiv2025

FACE: Few-shot Adapter with Cross-view Fusion for Cross-subject EEG Emotion Recognition

Cross-subject EEG emotion recognition is challenged by significant inter-subject variability and intricately entangled intra-subject variability. Existing works have primarily addressed these challenges through domain adaptation or generalization strategies. However, they typically require extensive target subject data or demonstrate limited generalization performance to unseen subjects. Recent few-shot learning paradigms attempt to address these limitations but often encounter catastrophic overfitting during subject-specific adaptation with limited samples. This article introduces the few-shot adapter with a cross-view fusion method called FACE for cross-subject EEG emotion recognition, which leverages dynamic multi-view fusion and effective subject-specific adaptation. Specifically, FACE incorporates a cross-view fusion module that dynamically integrates global brain connectivity with localized patterns via subject-specific fusion weights to provide complementary emotional information. Moreover, the few-shot adapter module is proposed to enable rapid adaptation for unseen subjects while reducing overfitting by enhancing adapter structures with meta-learning. Experimental results on three public EEG emotion recognition benchmarks demonstrate FACE's superior generalization performance over state-of-the-art methods. FACE provides a practical solution for cross-subject scenarios with limited labeled data.

eess.IV↗

arXiv2025

FDS: Frequency-Aware Denoising Score for Text-Guided Latent Diffusion Image Editing

Text-guided image editing using Text-to-Image (T2I) models often fails to yield satisfactory results, frequently introducing unintended modifications, such as the loss of local detail and color changes. In this paper, we analyze these failure cases and attribute them to the indiscriminate optimization across all frequency bands, even though only specific frequencies may require adjustment. To address this, we introduce a simple yet effective approach that enables the selective optimization of specific frequency bands within localized spatial regions for precise edits. Our method leverages wavelets to decompose images into different spatial resolutions across multiple frequency bands, enabling precise modifications at various levels of detail. To extend the applicability of our approach, we provide a comparative analysis of different frequency-domain techniques. Additionally, we extend our method to 3D texture editing by performing frequency decomposition on the triplane representation, enabling frequency-aware adjustments for 3D textures. Quantitative evaluations and user studies demonstrate the effectiveness of our method in producing high-quality and precise edits.

cs.CV↗

arXiv2025

Dynamic Topic Analysis in Academic Journals using Convex Non-negative Matrix Factorization Method

With the rapid advancement of large language models, academic topic identification and topic evolution analysis are crucial for enhancing AI's understanding capabilities. Dynamic topic analysis provides a powerful approach to capturing and understanding the temporal evolution of topics in large-scale datasets. This paper presents a two-stage dynamic topic analysis framework that incorporates convex optimization to improve topic consistency, sparsity, and interpretability. In Stage 1, a two-layer non-negative matrix factorization (NMF) model is employed to extract annual topics and identify key terms. In Stage 2, a convex optimization algorithm refines the dynamic topic structure using the convex NMF (cNMF) model, further enhancing topic integration and stability. Applying the proposed method to IEEE journal abstracts from 2004 to 2022 effectively identifies and quantifies emerging research topics, such as COVID-19 and digital twins. By optimizing sparsity differences in the clustering feature space between traditional and emerging research topics, the framework provides deeper insights into topic evolution and ranking analysis. Moreover, the NMF-cNMF model demonstrates superior stability in topic consistency. At sparsity levels of 0.4, 0.6, and 0.9, the proposed approach improves topic ranking stability by 24.51%, 56.60%, and 36.93%, respectively. The source code (to be open after publication) is available at https://github.com/meetyangyang/CDNMF.

cs.IR↗

arXiv2025

Mentor-Telemachus Bond: Transferring Knowledge in Semantic Communication via Contrastive Learning

Encoder, decoder and knowledge base are three major components for semantic communication. Recent advances have achieved significant progress in the encoder-decoder design. However, there remains a considerable gap in the construction and utilization of knowledge base, which plays important roles in establishing consensus among communication participants through knowledge transferring and sharing. Current knowledge base designs typically involve complex structures, which lead to significant computational overheads and heavy reliance on manually annotated datasets, making it difficult to adapt to existing encoder-decoder models. Hence, without knowledge transferring and sharing within the network results in poor generalization of encoder-decoder. This necessitates model training for specific tasks and datasets, significantly limiting the scalability of semantic communication systems to larger networks. To address these challenges, we propose an innovative Contrastive Representations Learning based Semantic Communication Framework (CRLSC). In CRLSC, the server-side pre-trained large model utilizes large-scale public datasets to construct shared knowledge base. Local-side encoders in terminal devices conduct training guided by shared knowledge base. These trained encoders can then build private knowledge bases from private datasets and fine-tune decoders for specific tasks. This simple and effective approach can facilitate the knowledge transferring across large-scale heterogeneous networks.

cs.NI↗

arXiv2025

DVMNet++: Rethinking Relative Pose Estimation for Unseen Objects

Determining the relative pose of a previously unseen object between two images is pivotal to the success of generalizable object pose estimation. Existing approaches typically predict 3D translation utilizing the ground-truth object bounding box and approximate 3D rotation with a large number of discrete hypotheses. This strategy makes unrealistic assumptions about the availability of ground truth and incurs a computationally expensive process of scoring each hypothesis at test time. By contrast, we rethink the problem of relative pose estimation for unseen objects by presenting a Deep Voxel Matching Network (DVMNet++). Our method computes the relative object pose in a single pass, eliminating the need for ground-truth object bounding boxes and rotation hypotheses. We achieve open-set object detection by leveraging image feature embedding and natural language understanding as reference. The detection result is then employed to approximate the translation parameters and crop the object from the query image. For rotation estimation, we map the two RGB images, i.e., reference and cropped query, to their respective voxelized 3D representations. The resulting voxels are passed through a rotation estimation module, which aligns the voxels and computes the rotation in an end-to-end fashion by solving a least-squares problem. To enhance robustness, we introduce a weighted closest voxel algorithm capable of mitigating the impact of noisy voxels. We conduct extensive experiments on the CO3D, Objaverse, LINEMOD, and LINEMOD-O datasets, demonstrating that our approach delivers more accurate relative pose estimates for novel objects at a lower computational cost compared to state-of-the-art methods. Our code is released at https://github.com/sailor-z/DVMNet/.

cs.CV↗

arXiv2025

Self-Ensembling Gaussian Splatting for Few-Shot Novel View Synthesis

3D Gaussian Splatting (3DGS) has demonstrated remarkable effectiveness in novel view synthesis (NVS). However, 3DGS tends to overfit when trained with sparse views, limiting its generalization to novel viewpoints. In this paper, we address this overfitting issue by introducing Self-Ensembling Gaussian Splatting (SE-GS). We achieve self-ensembling by incorporating an uncertainty-aware perturbation strategy during training. A $\mathbfΔ$-model and a $\mathbfΣ$-model are jointly trained on the available images. The $\mathbfΔ$-model is dynamically perturbed based on rendering uncertainty across training steps, generating diverse perturbed models with negligible computational overhead. Discrepancies between the $\mathbfΣ$-model and these perturbed models are minimized throughout training, forming a robust ensemble of 3DGS models. This ensemble, represented by the $\mathbfΣ$-model, is then used to generate novel-view images during inference. Experimental results on the LLFF, Mip-NeRF360, DTU, and MVImgNet datasets demonstrate that our approach enhances NVS quality under few-shot training conditions, outperforming existing state-of-the-art methods. The code is released at: https://sailor-z.github.io/projects/SEGS.html.

cs.CV↗

arXiv2025

Monte Carlo Diffusion for Generalizable Learning-Based RANSAC

Random Sample Consensus (RANSAC) is a fundamental approach for robustly estimating parametric models from noisy data. Existing learning-based RANSAC methods utilize deep learning to enhance the robustness of RANSAC against outliers. However, these approaches are trained and tested on the data generated by the same algorithms, leading to limited generalization to out-of-distribution data during inference. Therefore, in this paper, we introduce a novel diffusion-based paradigm that progressively injects noise into ground-truth data, simulating the noisy conditions for training learning-based RANSAC. To enhance data diversity, we incorporate Monte Carlo sampling into the diffusion paradigm, approximating diverse data distributions by introducing different types of randomness at multiple stages. We evaluate our approach in the context of feature matching through comprehensive experiments on the ScanNet and MegaDepth datasets. The experimental results demonstrate that our Monte Carlo diffusion mechanism significantly improves the generalization ability of learning-based RANSAC. We also develop extensive ablation studies that highlight the effectiveness of key components in our framework.

cs.CV↗

arXiv2025

Fully First-Order Methods for Decentralized Bilevel Optimization

This paper focuses on decentralized stochastic bilevel optimization (DSBO) where agents only communicate with their neighbors. We propose Decentralized Stochastic Gradient Descent and Ascent with Gradient Tracking (DSGDA-GT), a novel algorithm that only requires first-order oracles that are much cheaper than second-order oracles widely adopted in existing works. We further provide a finite-time convergence analysis showing that for $n$ agents collaboratively solving the DSBO problem, the sample complexity of finding an $ε$-stationary point in our algorithm is $\mathcal{O}(n^{-1}ε^{-7})$, which matches the currently best-known results of the single-agent counterpart with linear speedup. The numerical experiments demonstrate both the communication and training efficiency of our algorithm.

math.OC↗