VectorGym: A Multitask Benchmark for SVG Code Generation, Sketching, and Editing
为解决SVG生成、编辑等缺乏专业级基准问题，提出VectorGym，采用多任务强化学习方法优化，并提供人类标注数据和评估指标。
3 citationsRead paper
Selected work
Representative Papers
Albedo Estimation via Latent Bridge Matching
本文通过引入基于潜桥匹配的新架构解决了反照率估计中的物理一致性不足、计算成本高和泛化能力有限的问题。
0 citationsRead paper
Shedding Light: A Benchmark for Evaluating Lighting Understanding in Generative Image Models
本文提出了一种评估生成模型照明理解能力的基准，通过测试模型在真实照片中插入新对象时保持一致照明的能力，从而衡量其物理准确性。
0 citationsRead paper
MVTrack: Ultrafast Appearance-Free Moving Object Tracking from Compressed Bitstreams
This work addresses the high computational cost of existing RGB-based video object tracking methods, which hinders their large-scale deployment. The authors propose MVTrack, the first approach to achieve efficient tracking solely using motion vectors extracted from H.264 compressed bitstreams, entirely bypassing pixel-domain processing. MVTrack integrates a lightweight motion vector field detector (MVDet) with a minimalistic motion association module (MVLink) to enable accurate tracking without video decoding. Evaluated on the VIRAT dataset, MVTrack outperforms YOLOv2tiny in tracking accuracy while using 60× fewer parameters, requiring 40× lower FLOPs, and achieving 8.6× faster CPU inference speed, thereby significantly advancing the practicality and efficiency of compressed-domain tracking.
0 citationsRead paper
Superpixel-Based QUBO for Scalable Quantum-Enhanced Medical Image Segmentation
This work addresses the challenge of applying pixel-level QUBO (Quadratic Unconstrained Binary Optimization) models to medical image segmentation, which is hindered by the excessive number of variables incompatible with current quantum hardware. The authors propose a superpixel-based hierarchical modeling approach that integrates SLIC superpixels with QUBO for the first time. By constructing a Region Adjacency Graph (RAG), the segmentation task is reformulated as a low-dimensional QUBO optimization problem incorporating min-cut and smoothness constraints. This strategy substantially reduces variable count while preserving full-resolution image information and circumventing connectivity limitations of quantum annealers. Evaluated on the INbreast dataset, the method achieves a 97.3% reduction in problem size, a 33-fold acceleration in computation, and an average IoU of 0.76—improving accuracy by 4.2%—demonstrating a favorable balance among precision, efficiency, and scalability.
0 citationsRead paper
Recent publications
Latest Papers
Albedo Estimation via Latent Bridge Matching
本文通过引入基于潜桥匹配的新架构解决了反照率估计中的物理一致性不足、计算成本高和泛化能力有限的问题。
0 citationsRead paper
Shedding Light: A Benchmark for Evaluating Lighting Understanding in Generative Image Models
本文提出了一种评估生成模型照明理解能力的基准，通过测试模型在真实照片中插入新对象时保持一致照明的能力，从而衡量其物理准确性。
0 citationsRead paper
MVTrack: Ultrafast Appearance-Free Moving Object Tracking from Compressed Bitstreams
This work addresses the high computational cost of existing RGB-based video object tracking methods, which hinders their large-scale deployment. The authors propose MVTrack, the first approach to achieve efficient tracking solely using motion vectors extracted from H.264 compressed bitstreams, entirely bypassing pixel-domain processing. MVTrack integrates a lightweight motion vector field detector (MVDet) with a minimalistic motion association module (MVLink) to enable accurate tracking without video decoding. Evaluated on the VIRAT dataset, MVTrack outperforms YOLOv2tiny in tracking accuracy while using 60× fewer parameters, requiring 40× lower FLOPs, and achieving 8.6× faster CPU inference speed, thereby significantly advancing the practicality and efficiency of compressed-domain tracking.
0 citationsRead paper
Superpixel-Based QUBO for Scalable Quantum-Enhanced Medical Image Segmentation
This work addresses the challenge of applying pixel-level QUBO (Quadratic Unconstrained Binary Optimization) models to medical image segmentation, which is hindered by the excessive number of variables incompatible with current quantum hardware. The authors propose a superpixel-based hierarchical modeling approach that integrates SLIC superpixels with QUBO for the first time. By constructing a Region Adjacency Graph (RAG), the segmentation task is reformulated as a low-dimensional QUBO optimization problem incorporating min-cut and smoothness constraints. This strategy substantially reduces variable count while preserving full-resolution image information and circumventing connectivity limitations of quantum annealers. Evaluated on the INbreast dataset, the method achieves a 97.3% reduction in problem size, a 33-fold acceleration in computation, and an average IoU of 0.76—improving accuracy by 4.2%—demonstrating a favorable balance among precision, efficiency, and scalability.
0 citationsRead paper
Rethinking Expert Training for Model Merging with Prompt Learning
This work addresses the limitation of existing model merging methods, which typically assume that experts are obtained via full-parameter fine-tuning and thereby overlook how training strategies affect merge compatibility, often leading to inter-task interference. To mitigate this, the authors propose Dual-Tuned Experts (DTEs), a two-stage training strategy that first learns reusable prompts across tasks and then fine-tunes only the vision encoder while keeping the backbone frozen. This approach is the first to integrate prompt learning into expert training, decoupling prompt adaptation from backbone updates to substantially reduce task-specific perturbations. As a result, DTEs enhance both the stability and performance of merging heterogeneous experts. Experiments demonstrate that DTEs consistently outperform conventional fully fine-tuned experts across diverse CLIP architectures and expert types, while remaining compatible with mainstream model merging techniques.
0 citationsRead paper