Quantum Video Classification Leveraging Textual Video Representations
R Vinay, Badari Nath K · 2023
Quantum video classification is an emerging area of research which unites machine learning and quantum computing principles. This paper describes an approach for quantum video classification which combines video captioning and quantum natural language processing (QNLP) techniques. The proposed approach involves generating captions for videos and using these captions to produce quantum circuits, which are fed as inputs to a classifier running on a quantum simulator. This technique essentially reduces the quantum video classification problem into a quantum text classification problem. The text-based quantum circuits generated for videos also serve the purpose of acting as a new type of dataset for further research on quantum video processing. The proposed quantum video classifier performs well, with an accuracy of 89 percent when trained and tested on a subset of the kinetics dataset.