Fangda Chen

Fangda Chen, Shanshan Zhao, Longrong Yang et al.

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/.

Fangda Chen, Shanshan Zhao, Chuanfu Xu et al.

Recent advancements in customized video generation have led to significant improvements in the simultaneous adaptation of appearance and motion. While prior methods typically decouple appearance and motion training, the stage-wise strategy often introduces concept interference, resulting in inaccurate rendering of appearance features or motion patterns. Another challenge is appearance contamination, where background and foreground elements from reference videos distort the customized subject. In this work, we propose JointTuner, a novel framework that enables joint optimization of both appearance and motion components by leveraging two key innovations: Synaptic Low-Rank Adaptation (Synaptic LoRA) and Appearance-independent Temporal Loss (AiT Loss). Synaptic LoRA introduces a synaptic regulator, implemented as a context-aware linear activation layer, to dynamically guide LoRA modules to focus on either subject appearance or motion patterns, thereby enabling consistent optimization across spatial and temporal dimensions. AiT Loss disrupts the gradient flow of appearance-related components, guiding the model to focus exclusively on motion learning and minimizing appearance interference. JointTuner is compatible with both UNet-based models (e.g., ZeroScope) and Diffusion Transformer-based models (e.g., CogVideoX), supporting the generation of longer and higher-quality customized videos. Additionally, 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 can be found at https://fdchen24.github.io/JointTuner-Website.

Fangda Chen, Jintao Tang, Pancheng Wang et al.

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.

Fangda Chen, Jiacheng Nie, Lichuan Jiang et al.

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.