InCaRPose: In-Cabin Relative Camera Pose Estimation Model and Dataset

1. 1Collect a dataset of image pairs with varying camera poses
2. 2Train a Transformer-based model using frozen backbone features to predict relative poses
3. 3Fine-tune the model for specific in-cabin environments and lighting conditions
4. 4Evaluate the model's performance using metrics such as rotation and translation errors

Computer vision engineers and researchers on a team can benefit from InCaRPose for improving camera calibration, while automotive engineers can apply it to enhance in-cabin monitoring systems

💡 InCaRPose leverages frozen backbone features to improve robustness in highly distorted environments

🚗💻 InCaRPose: A new Transformer-based model for relative camera pose estimation in in-cabin automotive monitoring