Rendezvous: Attention mechanisms for the recognition of surgical action triplets in endoscopic videos.
Chinedu Innocent NwoyeTong YuCristians GonzalezBarbara SeeligerPietro MascagniDidier MutterJacques MarescauxNicolas PadoyPublished in: Medical Image Anal. (2022)
Keyphrases
- human actions
- action recognition
- recognizing actions
- human activities
- minimally invasive
- recognition rate
- endoscopic video
- action classification
- object recognition
- recognition accuracy
- activity recognition
- minimally invasive surgery
- motion history images
- video sequences
- pattern recognition
- human activity recognition
- recognition algorithm
- action detection
- intraoperative
- video frames
- feature extraction
- computer vision
- static images
- video content
- visual attention
- view invariant
- video analysis
- video data
- recognizing human actions
- recognition of human actions
- video database
- degrees of freedom
- robot assisted
- human object interactions
- reinforcement learning
- surgical instruments
- space time interest points