Abhijeet Solanki

AI Engineer

Edge AI Engineer

GenAI Engineer

IEEE-Published Researcher

Abhijeet Solanki

AI Engineer

Edge AI Engineer

GenAI Engineer

IEEE-Published Researcher

Download CV Contact Me

🏆 Accepted at IEEE Access:
"Survey of Navigational Perception Sensors' Security in Autonomous Vehicles."
📄 Submitted paper to NeurIPS 2025
🏆 Accepted at IEEE MWSCAS:
"Realistic GPS Spoofing via Customized CARLA GPS Navigation and Controller Systems"
🎤 Presented 3 papers at SouthEastCon 2025:
- "ReAL: Machine Learning Detection of Reflective Attacks Against Lidarometry"
- "GNAPing On the Job: Attacking and Defending Facial Detection on Edge Devices"
- "Towards Machine Learning Based Fingerprinting of Ultrasonic Sensors"
🧠 Accepted at FLAIRS 2025:
"Towards Trustworthy AI: Analyzing Model Uncertainty through MC Dropout and Noise Injection"
🚘 Accepted at ISCAS 2025:
"Mitigation of Camouflaged Adversarial Attacks in Autonomous Vehicles—A Case Study Using CARLA Simulator"

Adversarial Patch Removal for Stop Signs

Developed By: Abhijeet Solanki
Tech Stack: Python, PyTorch, Generative Restoration
Testbed: Quanser QCar2 (physical)
Result: Stop-sign detection 48% to 76%, zero clean-sign regressions
Github Url: https://github.com/ChiefAj23/qcar-patch-removal

A printed sticker on a stop sign can make a detector miss it in over half of frames. This generative restoration model removes the patch before detection on the QCar2 physical testbed, raising detection from 48% to 76% with zero regressions on clean signs.