Aashish Rai

Product Engineer

San Francisco, California, United States4 yrs 3 mos experience

Most Likely To Switch

Key Highlights

Expert in 3D/4D Computer Vision and Generative AI.
Developed innovative frameworks for multimodal generation.
Proven track record in photorealistic 3D face modeling.

Stackforce AI infers this person is a specialist in advanced computer vision and generative AI technologies.

Contact

aashish_rai@brown.edu LinkedIn

Skills

Core Skills

3d ReconstructionComputer Vision

Other Skills

3D Synthesis4D Reconstruction4D VisionAnacondaArtificial Neural NetworksC++Convolutional Neural Networks (CNN)Critical ThinkingDeep LearningDiffusion ModelsGenerative ModelsImage ProcessingIntel 8085JupyterLibreOffice

About

I am a CS Ph.D. candidate at Brown University, advised by Srinath Sridhar. My research focuses on the intersection of 3D/4D Computer Vision and Generative AI. I am currently leveraging Visual Language Models (VLMs) and novel Gaussian Splatting representations to learn and generate complex static and dynamic scenes. Research Highlights: 1. Gaussian Splatting & Neural Rendering: Developed UVGS, a 2D representation that enables direct application of image generative models to 3D Gaussian Splatting. I also proposed PackUV for efficient 4D volumetric video streaming. 2. Multimodal Generation: Created EgoSonics, a diffusion-based framework for generating synchronized audio for egocentric videos. 3. 3D Face Modeling: Previously at Carnegie Mellon University, I built SOTA frameworks for photorealistic 3D face generation and semantic editing.

Experience

Brown university

Doctoral Student

Aug 2023 – Present · 2 yrs 7 mos · Providence, Rhode Island, United States · On-site

4D ReconstructionNovel View SynthesisNeural Representation3D ReconstructionDiffusion ModelsProbabilistic Generative Models+3

Carnegie mellon university

Research Assistant

Sep 2021 – May 2023 · 1 yr 8 mos · Pittsburgh, Pennsylvania, United States · On-site

Worked with Human Sensing Lab in collaboration with Meta Reality Labs on 3D Face Reconstruction.

3D ReconstructionProbabilistic Generative Models3D SynthesisComputer Vision

Mcgill university

Undergraduate Research Intern

May 2020 – Oct 2020 · 5 mos · Montreal, Quebec, Canada

Research Internship with Shared Reality Lab, McGill University.
Worked on improving the latent space characterization of GANs for semantic face editing.

Probabilistic Generative Models

Norwegian university of science and technology (ntnu)

Undergraduate Student Researcher

Dec 2019 – Apr 2020 · 4 mos

Worked with Norwegian Biometrics Laboratory.
Designed and Implemented a lightweight CNN model to Generate 8x upscaled Super Resolution (SR) images from the 8x8, 16x16, and 24x24 LR images using a lightweight progressive residual propagating asymmetrical architecture with three modules (LF, HF feature extraction, and reconstruction).
The model was able to give appreciable results on benchmark datasets CelebA and LFW.

Probabilistic Generative Models

National institute of technology surat

Undergraduate Student Researcher

Jan 2019 – Nov 2019 · 10 mos · Surat

ML-CV Lab, Electronics Engineering Department
Implemented an Automated Attendance System using Deep Learning to mark the attendance of the entire class simultaneously and overcome the usual challenges of Occlusion, Orientation, and Luminance.
Transfer Learning was used to utilize the features of a pre-trained Inception model on a new dataset to achieve adequate results and to make the training computationally feasible.
The End-to-End accuracy was found to be 96.02 %

Probabilistic Generative Models