Akshay Raman
MSc in Computer Science | New York University, Courant Institute
Hi! I am Akshay Raman, a recent graduate in Computer Science from New York University. My research interests broadly span reinforcement learning, multimodal learning, and the scientific applications of AI.
I am currently a researcher at the AI4VS Lab at Columbia University, where I work under the supervision of Prof. Kaveri Thakoor. My current research focuses on multimodal learning approaches that integrate eye-tracking data for improved predictive performance. Previously, I had the opportunity to collaborate with researchers at the DICE Lab at NYU and the AI4Science Group at the University of Ottawa.
In my spare time, I enjoy music, running, and reading.
You can find my CV here.
Contact: ramanakshay112 [ at ] gmail [ dot ] com
News
| May 15, 2025 | 🎓 Graduated from New York University with a Master’s degree in Computer Science (AI Specialization). |
|---|---|
| Aug 18, 2023 | 🎓 Graduated from Vellore Institute of Technology with a Bachelor’s degree in Computer Science and Engineering. |
| Dec 20, 2022 |  Launched my personal website! |
Selected Projects
Finetuning Video Diffusion Models for Multi-view Consistency
Fine-tuned a video diffusion model to generate multi-view consistent object renderings from single-view inputs. Demonstrated that a curated high-quality 1% subset (10K objects) of the Objaverse dataset achieved performance comparable to full-scale training. (1M+ objects).
Canvas - A Template for Deep Learning Projects
Designed a flexible deep learning project template using pytorch and hydra. The template is based on the agent-environment interface in RL and supports all kinds of machine learning tasks.
Hierarchical CLIP-based Image Geolocation Prediction
Trained a CLIP-inspired image geolocation model that predicts the precise location of an image taken anywhere on earth. Designed a novel inference approach based on hierarchical feature clustering which achieves comparable performance while being ~100x more efficient than previous methods.
Continual Learning for Policy Gradient Methods
Masters Capstone Project
Developed novel incremental learning algorithms to train reinforcment learning agents on a variety of real-world environments. Modified batch-wise policy gradient methods using eligibility traces to eliminate data buffers, particularly for long horizon tasks.
Solving Optimal Transport using Deep Neural Networks
Project under Prof. Augusto Gerolin, Undergraduate Thesis
Developed gradient-based DNN appoximators to solve the optimal transport problem for high-dimensional data. Aimed to study application of OT in Density Functional Theory (DFT) to study dissociation of atoms.