You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is there a possible method to get the gt rotation matrix, translation vector of each image? i have realsense D435 camera, kinect DK v2. I notice we need these for custom model training. When record data from my D435 camera, i get the same rotation matrix, translation vector for each image even though i change the camera's angle,position or moving objects.
The text was updated successfully, but these errors were encountered:
Can you please elaborate? To finetune, you need to annotate data for real-world captures. Our repo doesn't provide any annotation pipelines for your data rather we use already provided NOCS data to fine-tune our model which is trained on synthetic data. Please also see this thread to get reasonable results captured from in-the-wild cameras. Hope it helps!
Is there a possible method to get the gt rotation matrix, translation vector of each image? i have realsense D435 camera, kinect DK v2. I notice we need these for custom model training. When record data from my D435 camera, i get the same rotation matrix, translation vector for each image even though i change the camera's angle,position or moving objects.
The text was updated successfully, but these errors were encountered: