A Nepalese Machine Learning (ML) Researcher Introduces ‘Papers-With-Video’ Browser Extension Which Allows Users To Access Videos Related To Research Papers On ArXiv

Amit Chaudhary, a machine learning (ML) researcher from Nepal, has recently introduced a browser extension that allows users to directly access videos related to research papers published on the platform arXiv.

ArXiv has become an essential resource for new machine learning (ML) papers. Initially, in 1991, it was launched as a storage site for physics preprints. In 2001 it was named ArXiv and had since been hosted by Cornell University. ArXiv has received close to 2 million submissions across various scientific research fields.

Amit obtained publicly released videos from 2020 ML conferences. He then indexed the videos and reverse-mapped them to the relevant arXiv links through pyarxiv, a dedicated wrapper for the arXiv API. The Google Chrome extension creates a video icon next to the paper title on the arXiv abstract page, enabling users to identify and access available videos related to the paper directly.

Many research teams are creating videos to accompany their papers. These videos can act as a guide by providing demo and other valuable information on the research document. In several situations, the videos are created as an alternative to traditional in-person presentations at AI conferences. This is useful in current circumstances as almost all panels have moved to virtual forms due to the Covid-19 pandemic.

The ‘Papers-With-Video’ extension enables direct video links for around 3.7k arXiv ML papers. Amit aims to figure out how to pair documents and videos related effectively but has different titles, and with this, he hopes to expand coverage to 8k videos. He has proposed community feedback and has now tweaked the extension’s functionality based on user remarks and suggestions.

The browser extension is not available on the Google Chrome Web Store yet. However, one can find the extension, installation guide, and further information on GitHub.

GitHub: https://github.com/amitness/papers-with-video

Paper List: https://gist.github.com/amitness/9e5ad24ab963785daca41e2c4cfa9a82

Tanushree Shenwai is a consulting intern at MarktechPost. She is currently pursuing her B.Tech from the Indian Institute of Technology(IIT), Bhubaneswar. She is a Data Science enthusiast and has a keen interest in the scope of application of artificial intelligence in various fields. She is passionate about exploring the new advancements in technologies and their real-life application.