I have deployed a object detection model using YOLOV10 and gradio on HF Spaces. I trained the model on a dataset called Rescuenet. This dataset includes drone imagery and my object detection model basically detects damaged and undamaged buildings. After that part, I did georeferencing based on images metadata like altitude, gimbal yaw degree, gimbal pitch degree. On original test images I use a tool called EXIFTool to get this data. Code works pretty fine on my local environment. In spaces, I can not obtain XMP data from images. Is there anyone who can help me about this issue? I assume the metadata is lost in the uploading phase. I used default values for the values I could not get from the image for now.
I am open to any suggestions or ideas to fix this issue.
Here is my space link: Rescuenet Damaged Building Detection - a Hugging Face Space by dolphinium
test images that you can use: test - Google Drive