The repo for the paper titled "A multimodal deep learning framework for scalablecontent based visual media retrieval"
Authors: Ambareesh Ravi, Amith Nandakumar
This research compares the performance of 3D CNNs for video retrieval to a novel framework that uses 2D CNNs + LSTMs for conjoint retrieval of images and videos.
https://arxiv.org/abs/2105.08665
The main codebase for the experiments in the paper is under training/