top of page
Damage Detection in Automobiles
Problem Statement: Given an image of a vehicle, detect all the damages such as dent, scratch, broken glass, crack, etc and report the extent of the damages detected.
Solution highlights:
-
Ensemble of deep learning algorithm that can detect, segment, and estimate the extent of the damage.
-
Deep learning APIs with a round trip time of 0.5 seconds per image.
-
Benchmarking Framework that produces detailed, quantitative report of the core algorithm accuracy.
Tech Stack used:
Image Quality Assessment, Super Resolution, Object Detection, Semantic Segmentation,
Instance Segmentation, Image Stitching, Homography, Edge Computing, Guided
Image Capture, PyTorch, Tensorflow, Deep Stream.
Previous
Next
Event Detection from ICU sensors
Damage Detection in Automobiles
bottom of page