Zhuoling Li

Automatic Method Illustration Generation for AI Scientific Papers via Drawing Middleware Creation, Evolution, and Orchestration

Zhuoling Li, Jiarui Zhang, Ping Hu, Jason Kuen, Jiuxiang Gu, Hossein Rahmani, Jun Liu

European Conference on Computer Vision (ECCV 2026)

• FigAgent uses reusable drawing middlewares and an Explore-and-Select strategy to generate high-fidelity, editable SVG method illustrations for AI papers.

HRDiT: Training-Free High-Resolution Image Generation with Off-the-Shelf Diffusion Transformer Models

Yu Xue, Haoxuan Qu, Zhuoling Li, Hongbin Xu, Jianxiong Yin, Simon See, Hossein Rahmani, Jun Liu

European Conference on Computer Vision (ECCV 2026)

• HRDiT enables training-free high-resolution image generation with off-the-shelf Diffusion Transformer models, without additional model training.

DiffGraph: An Automated Agent-driven Model Merging Framework for In-the-Wild Text-to-Image Generation

Zhuoling Li, Hossein Rahmani, Jiarui Zhang, Yu Xue, Majid Mirmehdi, Jason Kuen, Jiuxiang Gu, Jun Liu

IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2026)

• DiffGraph uses agents to automatically select and merge specialized text-to-image diffusion models for diverse, in-the-wild generation.

DiffIP: Representation Fingerprints for Robust IP Protection of Diffusion Models

Zhuoling Li, Haoxuan Qu, Jason Kuen, Jiuxiang Gu, Qiuhong Ke, Jun Liu, Hossein Rahmani

IEEE/CVF International Conference on Computer Vision (ICCV 2025)

• DiffIP protects diffusion-model IP by comparing robust representation fingerprints between victim and suspect models.

LongDiff: Training-free Long Video Generation in One Go

Zhuoling Li, Hossein Rahmani, Qiuhong Ke, Jun Liu

IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2025)

• LongDiff enables off-the-shelf short-video diffusion models to generate high-quality long videos in one pass, without training.

SceneLLM: Implicit Language Reasoning in LLM for Dynamic Scene Graph Generation

Hang Zhang, Zhuoling Li†, Jun Liu

Pattern Recognition (2025)

• SceneLLM maps video into implicit linguistic signals, enabling LLM-based reasoning for dynamic scene graph generation.

DisC-GS: Discontinuity-aware Gaussian Splatting

Haoxuan Qu*, Zhuoling Li*, Hossein Rahmani, Yujun Cai, Jun Liu

Annual Conference on Neural Information Processing Systems (NeurIPS 2024)

• DisC-GS models sharp boundaries with a discontinuity-aware rendering strategy, improving Gaussian Splatting near object edges.

FewVS: A Vision-Semantics Integration Framework for Few-shot Image Classification

Zhuoling Li, Yong Wang, Kaitong Li

ACM International Conference on Multimedia (ACM MM 2024)

• FewVS combines fine-grained LLM semantics with visual features and adaptively integrates both modalities for few-shot image classification.

Better Integrating Vision and Semantics for Improving Few-shot Classification

Zhuoling Li, Yong Wang

ACM International Conference on Multimedia (ACM MM 2023)

• BMI aligns semantic and visual latent spaces and uses prompt-based augmentation to improve few-shot image classification.