Worked on multiple projects with different research items across computer vision, differentiable filters, optimizations.

• Developed and refined Multi-object tracking (MOT) algorithms using Bayes Filter for Video Analytics for indoor and outdoor cameras.
• Developed monocular depth prediction models with varied advanced architecture-Vision Transformer (ViT) and multi-scale local planar guidance (LPG) blocks
• Developed multi-objective optimization technique base on Frank-Wolfe algorithm for training across multiple datasets
• Proposed Depth model shows 0.117, 0.416 on abs REL and RMS error metrics and 0.868 on d1 metric on NYU depth testset

Research topics: Differentiable Filters in Robot Learning and Human-robot Interaction. Monocular Depth Estimation Network for Human-robot and Human-environment interaction.

• Created a multimodal learning framework (α-MDF) using attention mechanism and differentiable filtering, which conducts state estimation in latent space with multiple modalities. Experimented on real-world tasks and validated the system on both rigid body robots and soft robots.
• Developed differentiable Ensemble Kalman Filters framework incorporating algorithmic priors for robot imitation learning, i.e., learning system dynamics from experience, missing observations, and learning representations in high-dimensional space
• Proposed framework shows 56.4% and 79.7% improvements in high-dimensional tasks, and 56.3% and 83.3% in low-dimension, on RMSE and MAE error metrics, comparing to the State-of-the-Art
• Implemented differentiable filters for HRI tasks, i.e., human-robot hugging, symbiotic walking

• Developed compact monocular depth prediction models and fine-tuned with temporal consistancy regularity
• Created a human gait (lower body) dataset associating depth data with biomechanical data
• Integrated depth prediction network with Interactive primitives for motor predictive modeling on an embedded system-Nvidia Jetson NX for real-time application