Efficient Segmentation: Learning Downsampling Near Semantic Boundaries
This addresses a critical bottleneck for automated processes like auto-piloting that rely on fast, accurate semantic segmentation.
The paper tackles the problem of semantic segmentation accuracy loss when downsampling input frames by proposing a content-adaptive downsampling technique that learns to sample near semantic boundaries. The method consistently outperforms uniform sampling in balancing accuracy and computational efficiency, improving boundary quality and reliability for small objects.
Many automated processes such as auto-piloting rely on a good semantic segmentation as a critical component. To speed up performance, it is common to downsample the input frame. However, this comes at the cost of missed small objects and reduced accuracy at semantic boundaries. To address this problem, we propose a new content-adaptive downsampling technique that learns to favor sampling locations near semantic boundaries of target classes. Cost-performance analysis shows that our method consistently outperforms the uniform sampling improving balance between accuracy and computational efficiency. Our adaptive sampling gives segmentation with better quality of boundaries and more reliable support for smaller-size objects.