Home About Publications Projects Education Work Skills
×

Kaustav Mukherjee

Computer Vision and Machine Learning

RESUME

My profile picture.
My profile picture.

Hi! I’m Kaustav. I’m currently an AI Engineer at Kaliber Labs in San Francisco, and a recent graduate of the CMU Robotics Institute, where I earned my Master's in Computer Vision.

My recent work spans from 3D vision and reconstruction pipelines to robotics and world models, with a sprinkle of AI conversational systems in between.

Publications

Surface Coverage Completeness for Quality-Assured Arthroscopic 3D Reconstruction

Dmitrii Bychkov, Rutwik Shah, Kaustav Mukherjee, Mitchell Ryan So

Accepted to ECCV 2026 Workshop DCAMI

Used truncated signed distance function (TSDF) fusion for real-time 3D reconstruction of the knee during arthroscopic surgery. We found that oblique viewing angles lead to significant drops in reconstruction quality, which has clinical relevance during surgery.

Paper ↗
2026
Projects
This image shows the planned camera trajectory along with the anchor trajectory, where the anchors are the targets the camera is looking at.

CMU Capstone: Cinematographic Camera Control

Developed an autonomous end-to-end cinematographic trajectory generation pipeline with hollywood production company Camera Control.

Poster ↗
2025
360 image deconstructed into cubemap projection.

Bundle Adjustment for 360° Pose Correction

Used bundle adjustment with neural image feature descriptors to improve gaussian splat 3D reconstruction quality from 360° images.

Report ↗
2025
TexTok-DiT Front Page Image

TexTok-DiT

Explored multimodal text-conditioned generation with a diffusion transformer architecture.

Report ↗ Code ↗
2025
AvatarsFTW Project Results

AvatarsFTW: 3D Human Avatars From the Wild

Created a pipeline to improve performance of SOTA 3D human avatar generation models on in-the-wild images.

Project Page ↗ Report ↗ Code ↗
2024
SmartResets Choose Policy Diagram

Smart Resets: Dynamic Schedules and Cost-Aware Resets for Faster Autonomous Reinforcement Learning

Developed a framework for improving exploration and safety in reinforcement learning by incorporating dynamic reset triggers and adaptive reset strategies.

Report ↗ Code ↗
2024
Image depicting airflow over a plane.

Neural Plane Optimization

Implementation and experimentation with neural plane optimization for 3D scene representation.

Code ↗
2024
Education

Carnegie Mellon University (CMU)

Master of Science in Computer Vision (MSCV)

Perception Research at CMU AirLab

2024–2025

National University of Singapore (NUS)

Bachelor of Engineering in Mechanical Engineering

Highest Distinction · 2nd Major in Innovation and Design

2020–2024
Work

Kaliber Labs

AI Engineer

2023–2026

F-Drones

Systems Engineering Intern

2022

SciFie Robotics

Robotics Engineering Intern

2021
Skills
Competencies
Generative AI Diffusion World Models LLM Fine-tuning Reinforcement Learning 3D Computer Vision Model Quantization Context Engineering
Frameworks
PyTorch TensorFlow NumPy OpenCV ROS2
Tools
Docker Redis NATS ONNX AWS Linux
Languages
Python C++ JavaScript
Kaustav Mukherjee Back to top ↑