PhD Student · University of Colorado Boulder

Pragyadipta
Adhya

Perceptual Quality · Image & Video · 3D Avatar Rendering · Chromatic Distortions

PhD student in the Department of Electrical, Computer and Energy Engineering at the University of Colorado Boulder, advised by the Great Prof. Alan C. Bovik. My research focuses on perceptual quality assessment of images, videos, and 3D avatar reconstruction and rendering — with a particular interest in chromatic distortions and how the human visual system perceives degradations in synthesised and rendered content. I conduct this research in collaboration with Meta Reality Labs.

01

About Me

Pragyadipta Adhya

I am a PhD student in the Department of Electrical, Computer and Energy Engineering at the University of Colorado Boulder, working under the supervision of Prof. Alan C. Bovik. My current research focuses on perceptual quality assessment of images, videos, and 3D avatar reconstruction and rendering, with a particular emphasis on chromatic distortions and how the human visual system responds to colour and quality degradations in synthesised content. This research is conducted in collaboration with Meta Reality Labs.

I hold an M.Tech in Signal Processing and Machine Learning from IIT Kharagpur, where my thesis developed a deep neural network-based codec for versatile image and video I-frame compression under the supervision of Dr. Debdoot Sheet. Prior to that, I completed a B.Tech in Electronics and Communication Engineering from Heritage Institute of Technology, Kolkata. My background spans convolutional neural networks, generative models, digital image processing, and learned compression.

I am broadly interested in understanding how machines and humans perceive visual quality — particularly in the context of emerging 3D content, avatar rendering pipelines, and the perceptual effects of colour and structural distortions.

Boulder, Colorado, USA
Perceptual Quality Assessment Image & Video Quality 3D Avatar Rendering Chromatic Distortions Deep Learning Learned Compression
02

Education

University of Colorado Boulder

PhD in Electrical, Computer and Energy Engineering

2025 — Present

Research Focus: Perceptual quality of images, videos & 3D avatar reconstruction/rendering; chromatic distortions; human visual perception

Indian Institute of Technology Kharagpur

M.Tech in Signal Processing and Machine Learning

2022 — 2024

Relevant Coursework:
  • Digital Image Processing
  • Deep Learning
  • ML for Signal Processing
  • Linear Algebra for Signals & Systems
  • Probability & Random Processes
Thesis: Deep Neural Network based Codec for Versatile Image and Video I-frame Compression

Heritage Institute of Technology, Kolkata

B.Tech in Electronics and Communication Engineering

2017 — 2021

Relevant Coursework:
  • Signals and Systems
  • Control Systems
  • Digital Signal Processing
  • Microprocessors & Microcontrollers
  • Object Oriented Programming
Thesis: Quantitative Analysis of Emotion using Color and Audio Stimuli

Certifications

NIELIT Kolkata

Certificate in Java & Oracle

2018 — 2019 · OOP with Java, MySQL

Add More Certifications

You can add your other certifications here (Coursera, Google, IBM, etc.)

03

Experience

Qualcomm India
Engineer in AI Camera Team · Hyderabad, India

July 2024 — July 2025

  • Deployed bit-exact deep neural model translation techniques, ensuring platform and pipeline independence.
  • Built a quantitative analysis pipeline for images, which reduces the time to quantify image quality on a large dataset from 23 hours to 3 hours.
  • Lead the development of an application to ensure the stability of deep neural models on edge devices.
Research Internship — IIT Kharagpur

January 2022 — June 2022

  • Developed flexible sensors for point-of-care diagnostics.
  • Meticulously mapped the sensor's temperature and pressure characteristics.
  • Integrated sensors with smart technology, enabling seamless monitoring via Bluetooth.
Ardent Computech Pvt. Ltd., Kolkata

May 2020 — June 2020

  • Led a team of three to build a robust insurance fraud detection system leveraging real-time claim data.
  • Benchmarked Decision Trees, Random Forest, and CNNs for classification effectiveness.
  • Delivered an enhanced fraud detection pipeline with improved precision and reliability.
Internship Programme — IIT Kanpur

June 2019 — July 2019

  • Devised an NLP tool merging PageRank Algorithm and DNNs for concise text summarisation.
  • Ensured adaptability to various document lengths while preserving essential content.
  • Significantly reduced email review time and elevated operational efficiency.
04

Research

Ongoing — PhD

Perceptual Quality Assessment of Images, Videos & 3D Avatar Rendering

Investigating how humans perceive quality degradations in synthesised and rendered visual content — including images, video sequences, and 3D avatar reconstruction pipelines. Research examines chromatic distortions, structural artefacts, and the role of the human visual system in evaluating rendered content quality.

Perceptual Quality 3D Avatar Rendering Chromatic Distortions Human Visual System Video Quality
PhD Research · University of Colorado Boulder · Advisor: Prof. Alan C. Bovik
Meta In collaboration with Meta Reality Labs
Completed — M.Tech Thesis

Deep Neural Network Codec for Versatile Image & Video I-frame Compression

Designed an end-to-end learned compression framework for versatile I-frame coding, outperforming classical codecs (HEVC/VVC) in rate-distortion performance. Explored hyperprior models, attention mechanisms, and learned entropy coding.

Learned Compression Generative Models Rate-Distortion PyTorch
M.Tech Thesis · IIT Kharagpur, 2022–2024
Completed — B.Tech Thesis

Quantitative Analysis of Emotion using Colour and Audio Stimuli

Investigated the correlation between emotional responses and multi-modal stimuli using signal processing and deep learning. Proposed a CNN-based framework for emotion quantification from colour and audio features.

Affective Computing Multimodal CNN Signal Processing
B.Tech Thesis · Heritage Institute of Technology
Completed

Flexible Sensor Development for Point-of-Care Diagnostics

Developed flexible sensors characterised for temperature and pressure, integrated with Bluetooth-enabled smart technology for remote health monitoring applications.

IoT Sensors Bluetooth Healthcare
Research Internship · IIT Kharagpur, 2022
Completed

Insurance Fraud Detection via Ensemble & Deep Learning

Built a fraud detection pipeline using Decision Trees, Random Forests, and CNNs on real insurance claim data. Benchmarked model performance across multiple classifiers with focus on precision and recall.

Machine Learning Ensemble Methods CNN Fraud Detection
Industry Internship · Ardent Computech, 2020
05

Software Projects

DVIC

Deep Neural Network Codec (DVIC)

Convolutional autoencoder for versatile image & video I-frame compression using channel ordinal learning and variable output resolution.

PyTorchCompressionAutoencoder
Image Deblurring

Convolutional Deblurring for Natural Images

Image restoration using Gaussian low-pass filtering, Point Spread Functions, edge deblurring, and convolution-based techniques.

Image ProcessingPSFOpenCV
Emotion Analysis

Emotion Analysis via Colour & Audio Stimuli

Quantitative framework using Decision Trees, Random Forest, and FCNNs to measure emotional response from multimodal inputs via FaceWeb.

Affective ComputingFCNNMultimodal
Text Summariser

Extractive Text Summariser

NLP pipeline merging Skip-gram word vectors, PageRank algorithm, and a FCNN to generate concise extractive document summaries.

NLPPageRankSkip-gram
Insurance Fraud

Insurance Fraud Detection

Real-time fraud detection pipeline on insurance claim data using Decision Trees, Random Forest, and CNNs benchmarked for precision and recall.

Random ForestCNNSklearn
06

Skills

Languages

C / C++

Frameworks & Libraries

Tools & Platforms

Hardware & Embedded

Bluetooth / IoT
07

Publications

1

Printable Graphene-Sustainable Elastomer-Based Cross Talk Free Sensor for Point of Care Diagnostics

Simran Sharma, Muthamil Selvan T, Susmita Naskar, Soumyadeep Mondal, Pragyadipta Adhya, et al., Tanmoy Mukhopadhyay, Titash Mondal

ACS Applied Materials & Interfaces

2

QoMEX 2026 Grand Challenge on Video Quality Assessment for Asymmetric Encoded Videos: Methods and Results

Jingwen Zhu, Yixu Chen, Hai Wei, Pierre Lebreton, Patrick Le Callet, Alexander Kopte, Amritha Premkumar, Anna Meyer, Baojun Li, Changsheng Gao, Christian Herglotz, Christian Timmerer, Dandan Zhu, Diwakara Reddy, Dong Liu, Dounia Hammou, Guangtao Zhai, Hadi Amirpour, Haoyun Cheng, Hichem Faraoun, Jonas Janzen, Krishna Srikar Durbha, Li Li, Marc Windsheimer, MohammadAli Hamidi, Mykyta Skipenko, Paul Wawerek-Lopez, Pragyadipta Adhya, Prajit T Rajendran, Rafał K Mantiuk, Shien Ke, Sid Ahmed Fezza, Simon Deniffel, Tao Jing, Wei Sun, Weixia Zhang, Xiangguang Chen, Yufan Liu, Zuowei Cao, Minhao Tang, Xiaoyan Sun, Xingwei Liu, Xiangyang Zhu, Yeganeh Chatri, Yenan Xu, Yifan Ding, Zijie Liu

2026 18th International Conference on Quality of Multimedia Experience (QoMEX)

3

PASTA — In Preparation

Pragyadipta Adhya, et al.

Expected submission — 2026

Coming Soon
08

Contact

Phone

+1 (720)-545-8034

Address

Boulder,
Colorado — 80303, USA

Social Profiles