University of York

Trinity College Dublin

Maharaja Surajmal Institute

Delhi Public School - R. K. Puram

Msc. in Computer Science - Augmented Reality

Worked on Computer Vision / Machine Vision solutions, including Object Classification, Object Detection, and Pose Estimation

• Designed custom neural network for object counting, outperforming other state-of-the-art 
deep learning solutions by up to 12%.

• Advanced the state-of-the-art for monocular visual odometry using deep learning and optical 
flow solutions, reducing drift error < 5%.

• Optimized SLAM solutions using Intel RealSense products.

• Synthesized datasets using Generative Adversarial Networks (GANs).

• Post-Training optimization with OpenVINO, quantization and sparsity.

• Embedded programming to optimize inference on Intel Movidius Vision Processor Units (VPU).

• Deployed multiple deep learning networks across different platforms using Docker and ROS.

• Collaborated with different partners for H2020 REFILLS (Project ID: 731590), a European 
research initiative.

 • Extensively worked with TensorFlow and ported the TensorFlow Object Detection pipeline to 
TensorFlow 2.0

 • Mentored interns with Python and Computer Vision.

 • Supported the Intel Edge AI incubator to promote deep learning on the edge and worked withstart-ups on using deep learning solutions with Intel Movidius products.

Tejas Pandey