
Gowthamaan Palani
Latest Updates
MDFN: Efficient Image Super-Resolution through Multi-Domain Feature Fusion
Multi-Domain Feature Network (MDFN), a novel and resource-efficient architecture for single image super-resolution that synergistically fuses features from three complementary domains: The Spatial Domain (using convolutions for local textures), the Multi-Scale Domain (using a Laplacian pyramid for hierarchical features), and the Frequency Domain (using Fourier transforms for global context).
📍 ICVGIP 2025
How I made my Blog website load in 0.3 seconds
Exploring Next.js 14 advanced features including SSR, SSG, ISR, and Image Optimization for optimal performance
Publications
Loading publications...
Education

Master of Science (MS) by Research, Engineering Design
Jan 2024 - Jun 2026
- Research Focus: Computer Vision and Medical Image Analysis
- Advisor: Dr. Ganapathy Krishnamurthi
- CGPA: 8.0/10

Bachelor of Technology (B.Tech), Electronics and Communication Engineering
2018 - 2022
- CGPA: 9.42/10
Experience

Project Officer
Jul 2026 - Present • Chennai, India
- Working on CPIS+, a multimodal deep learning framework augmenting the Clinical Pulmonary Infection Score (CPIS) for early prediction of Ventilator-Associated Pneumonia (VAP) before clinical onset

Research Assistant
Jan 2024 - Jun 2026 • Chennai, India
- Designed a novel, efficient multi-domain neural network for single-image super-resolution, combining deep learning with classical transforms (Laplacian, Fourier) to produce interpretable reconstructions
- Engineered an agentic AI system leveraging LLMs, retrieval-augmented generation, and knowledge graphs to structure free-form radiology reports with clinical grounding for diagnostic workflows
- Led end-to-end pre-processing and augmentation of large-scale, multi-modality medical imaging datasets to train and validate deep learning models with an emphasis on explainability

Software Engineer
Jun 2022 - Jun 2026 • Chennai & Puducherry, India
- Led full-stack development and launch of JeevanCloud, a remote patient-monitoring platform (Flutter, Django, WebSockets, Redis, PostgreSQL) for real-time data processing
- Built and deployed the IoT mobile application for the JeevanLite ICU ventilator in Flutter and native Android with an end-to-end CI/CD pipeline
- Optimized embedded firmware for the ventilator across STM32 and NRF microcontrollers, stabilizing performance across pressure, flow, oxygen, and mass sensor arrays

Software Development Intern
Oct 2020 - May 2022 • Puducherry, India
- Developed embedded-C algorithms for real-time flow-rate and volumetry estimation in mechanical ventilation using differential pressure sensors
- Designed a frame-based data protocol with flow control and error correction for robust BLE transfer between the ventilator hardware (NRF52832) and the mobile application
Skills
Have a project, an idea, or just want to say hello? - check my Standing Invitation