Abstract
Surgical gesture segmentation and recognition are important steps toward human-robot collaboration in robot-assisted surgery. In the human-robot collaboration paradigm, the robot needs to understand the surgeon's gestures to perform its tasks correctly. Therefore, training a computer vision model to segment and classify gestures in a surgery video is a focus in this field of research. In this paper, we propose a 2-phase surgical gesture recognition method and we evaluate empirically the method on JIGSAWS's suturing video dataset. Our method consists of a 3D convolutional neural network to detect the transition between 2 consecutive surgemes and a convolutional long short-term memory model for surgeme classification. To the best of our knowledge, ours is the first study aimed at detecting action transition in a multi-action video and to classify surgemes using an entire video portion rather than classifying individual frames. We also share our source code at https://github.comfiemiar/surgery-gesture-recog
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.