45 citations · 45 across the 2 of their papers we have counts for
2 papers
cs.CV2026
Can We Change "Stroke Size" to Simplify Flow Matching?
Yunwei Bai, Ying Kiat Tan, Yao Shu +1
Deep-learning-based image generation models are useful but difficult to train. Across the different time steps along a generative path, high-noise states present distinct optimizat…
cs.CV2011★ 45 cited
Towards Holistic Scene Understanding: Feedback Enabled Cascaded Classification Models
Congcong Li, Adarsh Kowdle, Ashutosh Saxena +1
Scene understanding includes many related sub-tasks, such as scene categorization, depth estimation, object detection, etc. Each of these sub-tasks is often notoriously hard, and s…