Akanimoh 'Sanmi Adeleye

Constantly learning and exploring.
Interested in Robotics, Computer Vision, and ML

Hello and welcome to my page

I am Akanimoh Adeleye though most people call me Sanmi. I am an NSF Graduate Research Fellow currently at the Contextual Robotics Institute at UC-San Diego.

I am a Computer Science PhD student in the Cognitive Robotics Lab. Before this, I spent a summer in NavLab at Carnegie Mellon University and with the robotics group at Oregon State University. I am excited by technological advancements and want to use robots to help people in need as well as push society forward

Outside of my research interests I also enjoy a slew of outdoor activities, getting lost in big cities, watching films, reading novels and slacklining between trees.

Community Education

CMU Robotics Ed
Kearny High School: STEM stories

Current and Past Work

My research focus is centered on leveraging contextual knowledge and geometric shape information to have robots efficiently organize everyday household objects. A few of my previous projects and experiences are shown below.

APPLE - Machine Learning Intern 2022

For this project I worked with a cross functional team to develop solutions for internal use at Apple. These solutions invloved Computer Vison and Machine learning based teniques as well as Data managment.

A. Adeleye, J. Hu, H. Christensen (2022). Putting away the Groceries with Precise Semantic Placements. 18th International Conference on Automation Science and Engineering (CASE).

J. Hu, A. Adeleye, H. I. Christensen (2022). Place-And-Pick-Based Re-Grasping Using Unstable Placement. ISRR - Robotics Research.
Link to Video

Washburn, A., Adeleye, A., An, T., and Riek, L.D. (2020). "Robot Errors in Proximate HRI: How Functionality Framing Affects Perceived Reliability and Trust". In ACM Transactions on Human Robot Interaction (THRI).

Intel - Software and Services Group 2017

For this project, I used OpenCV, Caffe and Intel’s Integrated Performance Primitives (IPP) to perform object recognition. The program took in a webcam feed or a set of images, performed region proposal and then used a deep learning model to classify objects. Using Intel's Vtune Amplifier, I analyzed run time speed comparing IPP and OpenCV for image resizing and region proposal. I then analyzed the run time speed of Caffe's classification comparing the MKL, ATLAS and OpenBlas math libraries.