A Review of Multi-modal and Multi-view Applications in Hand-Drawn Sketch Images
摘要
Hand-drawn sketches have played a significant role throughout human history, from the ancient world to the present, influencing everything from early childhood education to human psychology. While humans can effortlessly recognize and understand varying abstractions in these sketches, machines find these tasks challenging. Recent advancements in multi-modal models have opened new possibilities for computer vision to interpret hand-drawn sketches more effectively. This survey provides a comprehensive review of the application of multi-modal approaches in the domain of hand-drawn sketches, with a particular focus on generation and recognition tasks.