Hi,
I'm Sathish
Head - ADS
School of Innovation, KGiSL Institute of Technology, Coimbatore. Researcher in computer vision, deep learning and deepfake forensics, and a Python programmer at heart.
About
I'm Prof. Sathish Ramanujam
R. Sathish, M.E., (Ph.D.) - Head - ADS (Artificial Intelligence and Data Science) at the School of Innovation, KGiSL Institute of Technology, Coimbatore, and Head of the School of Programming Excellence.
I work in computer vision, deep learning and multimedia forensics, and have submitted my Ph.D. thesis in Machine Learning to Anna University. I build practical AI systems in Python, from face detection and deepfake video analysis to real-time pose estimation, and bring that hands-on work into teaching, student mentoring and technical events. Before teaching, I worked as a Malware Analyst at Dell.
- Ph.D., Machine Learning - Anna University (thesis submitted 2025-26)
- M.E., Software Engineering - Bannari Amman Institute of Technology, 2013 (CGPA 8.74, Distinction)
- B.Tech., Information Technology - Roever Engineering College, Perambalur, 2010
Skills
Technical Skills
I design, train and deploy computer vision and deep learning systems, tune them for accuracy and speed on limited hardware, and ship reusable code as open-source Python packages.
Python, SQL, Git, Jupyter, Linux, PyPI packaging
CNNs, Residual Attention Networks, spatio-temporal video models, GraphSAGE, hyperparameter optimization
OpenCV, MediaPipe (BlazePose), face detection, image quality assessment, object tracking, deepfake detection
TensorFlow, PyTorch, scikit-learn, NumPy, Pandas; reproducible experiments, latency and memory optimization
Research Interests
Explainable Deepfake Forensics Face Detection Computer Vision Deep Learning Spatio-Temporal Video Analysis Graph Neural Networks Surveillance AnalyticsProfessional Vision
To contribute to quality education, student innovation, industry-oriented learning and applied research in AI and Data Science by building reliable, efficient AI systems, sharing them as open tools, and guiding students to turn ideas into working products.
Research
Adaptive Hybrid Deep Learning Framework for Enhanced Deepfake Video Detection via Feature Integration
A three-part pipeline: an Enhanced Viola-Jones face detector, a hybrid spatio-temporal network (AHMcRAN) that learns manipulation patterns, and an improved training optimizer (ICLO). It outperformed CNN and ResNet baselines by 6.1% and 11.4% on Celeb-DF.
Journal Publications
An Intellectual Deep Fake Video Detection using Feature Integration Approach with Adaptive and Hybrid Deep Learning
R. Sathish, S. Daniel Madan Raja, E. Arul
KSII Transactions on Internet and Information Systems, 20(1), 327-353, Jan 2026
DOIConference Papers & Preprints
GraphSAGE for Real-Time Target Tracking in UAV Surveillance
IEEE ICITEICS 2025, 29 August 2025
IEEE XploreHybridFaceNet: A Model-Agnostic Validation Framework for Improving Precision in Face Detection Pipelines
MCT4SD 2025, Manila, Philippines - Springer Lecture Notes in Networks and Systems, May 2026
NeuralFit - AI Powered Fitness Application using MediaPipe with Pose Difference
Sathish R, Shivaprakash S, Sowmithran S, Sree Lekha S - 2025
DOI3 Registered Software Copyrights & PyPI Packages
Three software copyrights from my research, plus open-source Python packages on PyPI that implement its computer vision and ML components.
GitHubProjects
Deepfake Video Detection
AHMcRAN spatio-temporal network; +6.1% over CNN and +11.4% over ResNet on Celeb-DF.
Quality-Gated Face Detection
Viola-Jones + CNN verification with a quality gate: 97.2% detection at 23 FPS.
NeuralFit - AI Fitness Coach
Tracks 33 body landmarks to count reps and correct posture in under 80 ms per frame.
UAV Target Tracking
GraphSAGE for real-time target tracking in drone surveillance video.
Face Search
Search for a face in a video clip with AI and ML.
CCTV Footage Analyzer
Analyze CCTV footage in different aspects with AI and ML.
CV Receptionist
Identify faces and calculate the person count.
Air Quality Checker
Air quality analysis inside Volvo trucks using Raspberry Pi.
Smart Attendance System
Biometric-based attendance with Python.
Smart Trolley
Automatic bill calculation using an Intel Galileo board.
University Result Prediction
Predict university grades from statistical data.
Blind Reader
A smart reader for visually impaired people.
Experience
Head - ADS, School of Innovation
KGiSL Institute of Technology, Coimbatore
Head - School of Programming Excellence
KGiSL Institute of Technology, Coimbatore
Assistant Professor, AI & Data Science
KGiSL Institute of Technology, Coimbatore
Assistant Professor
Adithya Institute of Technology, Coimbatore
Assistant Professor
Shivani Group of Institutions
Malware Analyst
Dell Corporation Pvt. Ltd., Bangalore
11 Coding Events & Hackathons
- HackGenius - Smart India Hackathon 2025
- PyExpo 2026
- Institutional coding contests and hackathons
6 Invited Talks (2025-26)
- Dhanalakshmi College of Engineering, Chennai
- Phoenix Tech IT Services
- Papers presented at IEEE ICITEICS 2025 and Springer MCT4SD 2025, Manila
Professional Development
- PyCon India 2025
- 3 industry Faculty Development Programmes
- 7 seminars and workshops attended
Professor Sathish Academy
"It's not how much we give but how much love we put into giving."
Professor Sathish Academy (PSA) is a non-profit organisation offering training in programming, AI, IoT, web design and engineering subjects.
Embedded CC++C# .NET Core PythonMachine LearningIoT & Raspberry Pi Web DesignDigital MarketingEngineering SubjectsContact
Phone
+91 97914 06167Office
Dept. of AI & Data Science, KGiSL Institute of Technology,
365, Thudiyalur Road, Saravanampatti, Coimbatore - 641 035