arXiv subjects

Fangda Chen

Publications and source records attributed to Fangda Chen.

6 recordsLinked to original sources

arXiv2026

DART-SD: Diamond-topology Aware Retrieval and Tuning for Self-Distillation of Multi-Turn Tool-Calling Agents

Equipping Large Language Models (LLMs) with multi-turn tool-calling capabilities is essential for building autonomous agents. However, progress is fundamentally limited by the reliance on full-length trajectory imitation. For tasks involving multiple order-independent sub-goals, the optimal solution space forms a vast combinatorial diamond lattice. Forcing this rich topology into monolithic trajectories causes a severe topological collapse, indiscriminately penalizing valid alternative explorations and severely degrading policy diversity. To address this, we propose DART-SD (Diamond-topology Aware Retrieval and Tuning for Self-Distillation), a novel framework that shifts the paradigm from global forcing to topology-guided localized correction. DART-SD first models the execution process as a converging Interaction-State Transition Graph (ISTG), faithfully capturing the inherent diamond topology of successful and failed exploratory paths. During autonomous rollouts, the framework identifies the Critical Topological Breakpoint (CTB) and retrieves success-supported recovery references. Finally, we introduce a progressive self-distillation paradigm through CTB-guided localized supervision, ensuring that the training loss is calculated exclusively on the generated recovery steps while strictly protecting the valid reasoning prefix from destructive gradient updates. Experiments on complex multi-turn tool-calling benchmarks demonstrate that DART-SD significantly outperforms traditional full-trajectory baselines.

cs.CL↗

arXiv2026

FreeSpec: Training-Free Long Video Generation via Singular-Spectrum Reconstruction

Video diffusion models perform well in short-video synthesis, but their training-free extension to long videos often suffers from content drift, temporal inconsistency, and over-smoothed dynamics. Existing methods improve temporal consistency by combining a global branch with a local branch, but they often further decompose appearance consistency and temporal dynamics within each branch using predefined criteria. This assignment is unreliable when appearance and action progression are tightly coupled, such as in camera motion and sequential motion. We analyze the video temporal extension issue from a singular-spectrum perspective and show that enlarged self-attention windows induce spectral concentration: spectral energy becomes dominated by a few low-rank singular directions, preserving coarse structure but suppressing high-rank spatial details and motion-rich temporal variations. To mitigate this problem, we propose FreeSpec, a training-free spectral reconstruction framework for long-video generation. FreeSpec decomposes global and local features with singular value decomposition, and uses the global branch as low-rank spectral guidance and the local branch as a high-rank reconstruction basis. This spectrum-level fusion avoids the rigid feature partitioning of previous decomposition rules, preserving long-range consistency while better retaining spatial details and temporal dynamics. Experiments on Wan2.1 and LTX-Video demonstrate that FreeSpec improves long-video generation, especially for temporal dynamics, while maintaining strong visual quality and temporal consistency. Project demo: https://fdchen24.github.io/FreeSpec-Website/.

cs.CV↗

arXiv2026

SWGCN: Synergy Weighted Graph Convolutional Network for Multi-Behavior Recommendation

Multi-behavior recommendation paradigms have emerged to capture diverse user activities, forecasting primary conversions (e.g., purchases) by leveraging secondary signals like browsing history. However, current graph-based methods often overlook cross-behavioral synergistic signals and fine-grained intensity of individual actions. Motivated by the need to overcome these shortcomings, we introduce Synergy Weighted Graph Convolutional Network (SWGCN). SWGCN introduces two novel components: a Target Preference Weigher, which adaptively assigns weights to user-item interactions within each behavior, and a Synergy Alignment Task, which guides its training by leveraging an Auxiliary Preference Valuator. This task prioritizes interactions from synergistic signals that more accurately reflect user preferences. The performance of our model is rigorously evaluated through comprehensive tests on three open-source datasets, specifically Taobao, IJCAI, and Beibei. On the Taobao dataset, SWGCN yields relative gains of 112.49% and 156.36% in terms of Hit Ratio (HR) and Normalized Discounted Cumulative Gain (NDCG), respectively. It also yields consistent gains on IJCAI and Beibei, confirming its robustness and generalizability across various datasets. Our implementation is open-sourced and can be accessed via https://github.com/FangdChen/SWGCN.

cs.IR↗

arXiv2025

DEAP-3DSAM: Decoder Enhanced and Auto Prompt SAM for 3D Medical Image Segmentation

The Segment Anything Model (SAM) has recently demonstrated significant potential in medical image segmentation. Although SAM is primarily trained on 2D images, attempts have been made to apply it to 3D medical image segmentation. However, the pseudo 3D processing used to adapt SAM results in spatial feature loss, limiting its performance. Additionally, most SAM-based methods still rely on manual prompts, which are challenging to implement in real-world scenarios and require extensive external expert knowledge. To address these limitations, we introduce the Decoder Enhanced and Auto Prompt SAM (DEAP-3DSAM) to tackle these limitations. Specifically, we propose a Feature Enhanced Decoder that fuses the original image features with rich and detailed spatial information to enhance spatial features. We also design a Dual Attention Prompter to automatically obtain prompt information through Spatial Attention and Channel Attention. We conduct comprehensive experiments on four public abdominal tumor segmentation datasets. The results indicate that our DEAP-3DSAM achieves state-of-the-art performance in 3D image segmentation, outperforming or matching existing manual prompt methods. Furthermore, both quantitative and qualitative ablation studies confirm the effectiveness of our proposed modules.

cs.CV↗

arXiv2025

JointTuner: Appearance-Motion Adaptive Joint Training for Customized Video Generation

Recent advancements in customized video generation have led to significant improvements in the simultaneous adaptation of appearance and motion. Typically, decoupling the appearance and motion training, prior methods often introduce concept interference, resulting in inaccurate rendering of appearance features or motion patterns. In addition, these methods often suffer from appearance contamination, in which background and foreground elements from reference videos distort the customized video. This paper aims to alleviate these issues by proposing JointTuner. The core motivation of our JointTuner is to enable joint optimization of both appearance and motion components, upon which two key innovations are developed, i.e., Gated Low-Rank Adaptation (GLoRA) and Appearance-independent Temporal Loss (AiT Loss). Specifically, GLoRA uses a context-aware activation layer, analogous to a gating regulator, to dynamically steer LoRA modules toward learning either appearance or motion while maintaining spatio-temporal consistency. Moreover, with the finding that channel-temporal shift noise suppresses appearance-related low-frequencies while enhancing motion-related high-frequencies, we designed the AiT Loss. This loss adds the same shift to the diffusion model's predicted noise during fine-tuning, forcing the model to prioritize learning motion patterns. JointTuner's architecture-agnostic design supports both UNet (e.g., ZeroScope) and Diffusion Transformer (e.g., CogVideoX) backbones, ensuring its customization capabilities scale with the evolution of foundational video models. Furthermore, we present a systematic evaluation framework for appearance-motion combined customization, covering 90 combinations evaluated along four critical dimensions: semantic alignment, motion dynamism, temporal consistency, and perceptual quality. Our project homepage is available online.

cs.CV↗

arXiv2024

Few-shot Calligraphy Style Learning

We introduced "Presidifussion," a novel approach to learning and replicating the unique style of calligraphy of President Xu, using a pretrained diffusion model adapted through a two-stage training process. Initially, our model is pretrained on a diverse dataset containing works from various calligraphers. This is followed by fine-tuning on a smaller, specialized dataset of President Xu's calligraphy, comprising just under 200 images. Our method introduces innovative techniques of font image conditioning and stroke information conditioning, enabling the model to capture the intricate structural elements of Chinese characters. The effectiveness of our approach is demonstrated through a comparison with traditional methods like zi2zi and CalliGAN, with our model achieving comparable performance using significantly smaller datasets and reduced computational resources. This work not only presents a breakthrough in the digital preservation of calligraphic art but also sets a new standard for data-efficient generative modeling in the domain of cultural heritage digitization.

cs.CV↗