Automated Surgical Activity Recognition with One Labeled Sequence
This work addresses the tedious and expensive manual annotation process in robot-assisted surgery, though it is incremental as it builds on prior activity recognition methods.
The paper tackles the problem of automated surgical activity recognition with scarce annotations, demonstrating feasibility using only one labeled sequence for training and showing that unsupervised representation learning significantly improves performance.
Prior work has demonstrated the feasibility of automated activity recognition in robot-assisted surgery from motion data. However, these efforts have assumed the availability of a large number of densely-annotated sequences, which must be provided manually by experts. This process is tedious, expensive, and error-prone. In this paper, we present the first analysis under the assumption of scarce annotations, where as little as one annotated sequence is available for training. We demonstrate feasibility of automated recognition in this challenging setting, and we show that learning representations in an unsupervised fashion, before the recognition phase, leads to significant gains in performance. In addition, our paper poses a new challenge to the community: how much further can we push performance in this important yet relatively unexplored regime?