Work Experience

Position: Graduate Research Assistant

Employment Type: Full Time

Location: West Lafayette, Indiana, USA

Responsibilities:

• Advised by Dr. Amy R. Reibman
• Understanding, evaluating, and mitigating the impact of video compression of computer vision systems
• Developing task-aware video compression algorithms that improve system performance on compressed video data
• Designed image quality estimators and evaluation protocols to correlate the performance of deep learning models to image quality
• Research published in IEEE ICIP, IEEE MMSP, IEEE MIPR, and Electronic Imaging

Position: Machine Learning and Video Coding Intern

Employment Type: Fall internship

Location: Sunnyvale, CA, USA

Responsibilities:

• Advised by Dr. Peng Yin and Dr. Tong Shao
• Designed an end-to-end neural network based compression algorithm for 3D Gaussian Splats
• Utilized traditional video codecs to exploit inter-frame redundancies for 3DGS compression
• Worked on post-training model pruning and quantization methods, shifting focus from modality-specific compression
• Research submitted to IEEE ICIP

Position: Machine Learning and Video Coding Intern

Employment Type: Summer internship

Location: Sunnyvale, CA, USA

Responsibilities:

• Advised by Dr. Guan-Ming Su and Dr. Anustup Choudhury
• Mastered design principles of Neural Field architectures for images, video, and audio signals, encompassing Fourier embeddings, Sine/Swish activations, and Convolution-based structures (NeRV, NeRF)
• Designed a novel implicit Neural Field architectures for joint audio-video representation using time-stamp coordinates (NeRVA)
• Worked on post-training model pruning and quantization methods, shifting focus from modality-specific compression
• Patent submitted and research published in IEEE ICME

Position: Video Coding Intern

Employment Type: Summer internship

Location: Cupertino, CA, USA

Responsibilities:

• Advised by Dr. Alexis Tourapis
• Analysis of Common Test Conditions (CTC) sequences in terms of spatial, temporal, noise and compressibility characteristics
• Utilized ML techniques such as super-resolution and temporal resolution for Scalable Video Compression

Position: Research Assistant

Employment Type: Full Time

Location: Bengaluru, KA, India

Responsibilities:

• Advised by Dr. Abhay Sharma
• Designed a real-time front-end Web Server System that delivers live RTMP and HLS video streams. Integrated features such as content sharing, discovering, routing, congestion managing, and load balancing
• Colloborated on implementing solutions tackling traffic analytics such as vehicle count, detection of license-plates, speed computation, and estimation of queue-length
• The full framework has been successfully installed and is operational in Electronic City, Bangalore, KA, India

Position: Self-driving Engineer

Employment Type: Full Time

Location: Bengaluru, KA, India

Responsibilities:

• Pedestrian and lane detection to achieve Level 3 autonomy for trucks in a controlled environment. Also, worked on the Over-The-Air Update infrastructure

Position: R&D Software Engineer

Employment Type: Full Time

Location: Bengaluru, KA, India

Responsibilities:

• Automation frameworks and testbeds for involved optical networks. Worked on aspects of Add-Drop Multiplexers, Network Distribution Cards and Ethernet Layer 2 functionalities