This is an LWIR stereo-hyperspectral database to develop HADAR algorithms for thermal navigation. Based on this database, one can develop algorithms for TeX decomposition to generate TeX vision. One can also develop algorithms about object detection, semantic or scene segmentation, optical or scene flow, stereo depth etc. based on TeX vision instead of traditional RGB or thermal vision.
MIT License
164
stars
26
forks
source link
About the excessively low mIoU output from visualization.py #25
Hello, I encountered an issue while training my model where it was missing the emiLib.mat file. However, I found the file under the TxNet folder and placed it in the HADAR_database folder. After that, I used preprocess_data.py to obtain the .npy files. But when I used the visualization.py file later, the output mIoU was very low. I set the number of epochs to 40000, but I noticed that the val_iou reached 0.17 around epoch 2000 and hardly improved during the subsequent training. I would like to ask what could be the possible reasons for this?
Hello, I encountered an issue while training my model where it was missing the emiLib.mat file. However, I found the file under the TxNet folder and placed it in the HADAR_database folder. After that, I used preprocess_data.py to obtain the .npy files. But when I used the visualization.py file later, the output mIoU was very low. I set the number of epochs to 40000, but I noticed that the val_iou reached 0.17 around epoch 2000 and hardly improved during the subsequent training. I would like to ask what could be the possible reasons for this?![屏幕截图 2024-05-13 192637](https://github.com/FanglinBao/HADAR/assets/139959653/74e7157d-a29a-4279-bd62-1ead5f054cf1)