Depth-Categorizing GPU-Accelerated Deep Neural Networks Perform Fast Semantic Segmentation of RGB-D Scenes

Interesting for computer vision and animation, the paper by Nico Höft, Hannes Schulz, and Sven Behnke, “Fast Semantic Segmentation of RGB-D Scenes with GPU-Accelerated Deep Neural Networks” categorizes the surface to which each pixel in an image belongs. Semantic scene segmentation is a major challenge on the way to functional computer vision systems that can separately label the objects in a scene. The authors report their method can reach competitive performance at a high frame rate on the NYU Depth V2 dataset.

Labeled Image from the NYU dataset (Click to visit site) [citation Indoor Segmentation and Support Inference from RGBD Images ECCV 2012]

For more about deep-learning see:

• Deep-learning Webinar Demonstrates Handwriting Recognition and Efforts to Teach Drone to Fly Down a Wooded Path

• IBM TrueNorth a “Bee Brain” on a SyNAPSE Chip That Uses 70 mW

• Deep-learning Teaching Code Achieves 13 PF/s on the ORNL Titan Supercomputer

Click here for more TechEnablement machine-learning articles and tutorials!