I’m new to computer vision and designing a program for car damage detection system on Google Colab. I had trained two models for detecting damage and car parts as instance segmentation done using Detectron2, which I’m trying to load in the system but it keeps failing due to “too many indices for tensor of dimesion 3” error despite the input shape being correct. Please help me out.
Extra info: The input consists of 8 images of standard angles of a damaged car and result would be the car parts which are detected as damaged.