Mustafa Shaikh

I work with deep learning, computer vision, and robotics.

I'm currently working on a multi-view 3D pose estimation problem at a neuroscience lab at the Salk Institute. I recently wrapped up a project on multi-object tracking for animal and cell video data using deep learning. The project is fully open source - check it out at dreem.sleap.ai.

During my master's research, I worked on a trajectory optimization problem in a pursuer-evader scenario. We demonstrated accurate and collision free tracking on a real robot.

Previously, I worked in industry for ~5 years in large organizations in a data science role. I worked closely with business stakeholders and led projects end to end.

Check out some of my other projects below!

Resume

Publications

Control Strategies for Pursuit-Evasion Under Occlusion Using Visibility and Safety Barrier Functions

Minnan Zhou*, Mustafa Shaikh*, Vatsalya Chaubey*, Patrick Haggerty, Shumon Koga, Dimitra Panagou, Nikolay Atanasov
*Equal contribution

Accepted to the IEEE International Conference on Robotics and Automation (ICRA) 2025
project page / arXiv

Projects

	LLM Workflow - Research Paper Retrieval An LLM workflow with a finetuned Llama 3.2 1B Instruct that retrieves relevant research papers from Arxiv based on user chat queries. Integrated with Cursor via MCP.
	Visual Inertial SLAM Landmark and trajectory estimation from calibrated stereo camera features and IMU measurements.
	2D Occupancy and Texture Mapping Particle filter for occupancy and texture mapping using LiDAR, RGBD camera, and IMU data.
	3D Motion Planning Classical 3D motion planning with obstacles.
	(C++) String Library with Buffer Manager A barebones String library with a buffer manager.
	(C++) Archive Tool A general purpose archive tool with compression, add, remove, and retrieve functionality.
	(C++) In-memory JSON Database and Query System An in-memory JSON model with a query language to select, filter, and aggregate data.

Website template from Jon Barron