Thermal Facial Landmarks Detection
A leading manufacturing firm specializing in high-volume production of automotive components. The company operates multiple production lines and faces frequent scheduling bottlenecks due to fluctuating customer demands.
Detecting facial landmarks in thermal images is challenging due to variations in image quality, lack of visible light features, and limited publicly available datasets. Accurate detection is essential for applications in security, healthcare, and human-computer interaction
Developed and optimized a facial landmark detection pipeline tailored for thermal imaging applications.
Established a robust baseline model, validated through industry-standard evaluation metrics to ensure high accuracy and reliability.
Engineered custom tools for facial feature extraction and enhanced model resilience across low-quality image and video inputs.
The trained model demonstrated reliable landmark detection on high-quality thermal images but showed performance variations on lower-quality inputs. The study highlighted key challenges in thermal-based facial recognition and the need for further optimization.
