Hello there!

I'm Kshitij Ambilduke

I am currently pursuing my Master's in Computer Science (AI-track) at Paris-Saclay University. Prior to this, I worked in the SARDINE lab as a Trainee Researcher at the Instituto de Telecomunicações, Lisbon. I finished my bachelor's in Electronics and Communication Engineering at Visvesvaraya National Institute of Technology (VNIT), India in 2023.

Previously, I pursued a research internship co-advised by Dr. Andre Martins and Dr. Bruno Martins at Instituto Superior Técnico, Lisbon where worked on Visual Question answering. I am interested in Deep Learning and Natural Language Processing and I'm particularly fascinated by the challenges lying in the intersection of language with other modalities.

Besides this, I play badminton, cricket, and Valorant and listen to pop music.

Here's a big thumbs up for visiting me : )

Projects

For a more detailed description, please refer to the projects page

Enhancing context through contrast

We proposed an additional novel pre-training step based on Barlow twins contrastive loss for Neural Machine Translation.

Medical Visual Question answering

Reviewed & compiled notes on several research papers on VQA and implemented several Medical VQA baselines using Meta AI's MMF framework

Neural Machine Translation

Implemented several NMT models from scratch in PyTorch including SOTA models like Transformers

Language Modelling

Implemented several word and character level language models like RNN, LSTM & GRU in PyTorch.

Image Captioning

Studied and implemented several image captioning models using PyTorch. Further, wrote a review paper on the basis of the studies model.

Visual Transformers

Reviewed several works on transformer-based vision models and implemented Visual Transformers from scratch in PyTorch.

Memes-vs-notes

Trained and tested DNNs and CNNs, built from scratch in NumPy and PyTorch respectively, on a custom Memes-vs-notes dataset.

News

[May 2024] Awarded with the France Excellence Charpak Master Scholarship 2024 for pursuing master's at Paris-Saclay University.
[Jul 2023] Received a competitive scholarship worth €1500 to attend the 13th edition of LxLMS: Lisbon Machine Learning School
[Feb 2023] My paper titled "Attending to Transforms: A Survey on Transformer-based Image Captioning" was accepted at PCEMS 2023
[Oct 2021] My paper titled "Enhancing Context Through Contrast" was accepted at the NeurIPS 2022 Pre-registration workshop.
[Sep 2021] Ranked amongst the top 100 teams in the Amazon ML Challenge 2021.
[Aug 2021] Elected as the Chairperson for IEEE VNIT Student Branch.

Get in touch : kshitij.a.vnit@gmail.com