Holistic Surgical Phase Recognition with Hierarchical Input Dependent State Space Models

Surgical workflow analysis is essential in robot-assisted surgeries, yet the long duration of such procedures poses significant challenges for comprehensive video analysis. Recent approaches have predominantly relied on transformer models; however, their quadratic attention mechanism restricts efficient processing of lengthy surgical videos. In this paper, we propose a novel hierarchical input-dependent state space model that leverages the linear scaling property of state space models to enable decision making on full-length videos while capturing both local and global dynamics. Our framework incorporates a temporally consistent visual feature extractor, which appends a state space model head to a visual feature extractor to propagate temporal information. The proposed model consists of two key modules: a local-aggregation state space model block that effectively captures intricate local dynamics, and a global-relation state space model block that models temporal dependencies across the entire video. The model is trained using a hybrid discrete-continuous supervision strategy, where both signals of discrete phase labels and continuous phase progresses are propagated through the network. Experiments have shown that our method outperforms the current state-of-the-art methods by a large margin (+2.8% on Cholec80, +4.3% on MICCAI2016, and +12.9% on Heichole datasets). Code will be publicly available after paper acceptance.

mm-Wave Radar Hand Shape Classification Using Deformable Transformers

A novel, real-time, mm-Wave radar-based static hand shape classification algorithm and implementation are proposed. The method finds several applications in low cost and privacy sensitive touchless control technology using 60 Ghz radar as the sensor input. As opposed to prior Range-Doppler image based 2D classification solutions, our method converts raw radar data to 3D sparse cartesian point clouds.The demonstrated 3D radar neural network model using deformable transformers significantly surpasses the performance results set by prior methods which either utilize custom signal processing or apply generic convolutional techniques on Range-Doppler FFT images. Experiments are performed on an internally collected dataset using an off-the-shelf radar sensor.