• Researching High Performance Computing for 3D Scene Representations to accelerate the Gaussian Splatting pipeline using C/CUDA.

• Working at the Human Sensing Lab, directed by Prof. Fernando De la Torre.

Reduced training epoch times by 49% by developing an accelerated neural network pipeline.

Attained a 6x speedup by implementing L1 cache tiling, minimizing memory access latency in Fused Multiply-Add kernels.

Developed an optimized Softmax kernel that reduced compute time by 2 orders of magnitudes through shared memory reduction and coalesced accesses.

Executed the Neural Radiance Fields research methodology to deliver novel view synthesis and 3D reconstruction.

Achieved an average Peak Signal-to-Noise Ratio score of 29.20 on the test set for View Synthesis.

Leveraged Phong Illumination along with reflection and shadows to accurately simulate light physics.

Implemented efficient ray-geometry intersection algorithms for 3D primitives at arbitary orientations.

Utilized the SIFT algorithm to extract robust keypoints and descriptors, combined with RANSAC for estimating geometric transformations between image pairs.

Introduced weighted blending techniques to ensure smooth transitions. Reduced artifacts by 90% and overall +32% in image alignment.

Implemented the ViT-B/16 model from the research paper ”An image is worth 16x16 words” in PyTorch.

Authored a DataX Journal article analyzing the transformative impact of Transformer Architecture in the 2020.

• Advanced a suite of deep learning projects, sharpening expertise in 3D/computer vision, and graphics domains.

• Fostered lab growth and collaboration by recruiting new associates and engaging in 20+ paper discussions, seminars, and talks to stay current with emerging research.

Co-led with 2 research supervisors to guide and mentor members on projects and initiatives.

Directed a project blocking 96% of explicit content to enhance the security of children in online environments.

Architected a queue-based parallel computing framework, accelerating machine learning processing by 3x across diverse content channels, including text, image, speech, and video.

• Relevant Coursework: Calculus and Linear Algebra — Advance Calculus and Complex Analysis — Probability and Statistics — Discrete Mathematics — Digital Image Processing — DSA — Design and Analysis of Algorithms — Advanved OOP — Operating Systems — DB Management Systems — ANNs — ML for Data Analysis — Compiler Design