Sketch Research

Sketch research focuses on bridging the gap between human-drawn sketches and computer vision, aiming to improve how computers understand and utilize sketched information. Current research emphasizes developing robust models that handle sketches of varying quality and detail, from quick, impressionistic drawings to precise, detailed renderings, often employing diffusion models and optimal transport algorithms to achieve this. This work is significant for advancing image generation, inpainting, retrieval, and other visual understanding tasks, as well as providing insights into human-computer interaction and the nature of visual communication. The development of standardized benchmark datasets and evaluation metrics is also a key focus, improving the rigor and comparability of research findings.