Spaces:
Sleeping
Apply for community grant: Academic project (gpu)
Hello Sir,
I and my colaborator are senior students at Ho Chi Minh University of Science in Vietnam. We are currently working on graduation thesis, which focuses on object detection using fisheye camera based on Fisheye8K dataset. Our study has yielded promising results through the utilization of a custom-designed YOLOv9 model and loss function for fisheye camera. Notably, our model has outperformed performance compared to the YOLOv9-e base model and other SOTA one-stage real-time models. Our loss function also help the model increase the mAP@50:95 by 2.3% and mAP@50 by 2.9%.
We kindly ask for your assistance in providing GPU to our space for demo purpose during our graduation thesis defence and also during testing before showing it.
- yolov9-e trained on Fisheye8K result:
- Our model trained on image.png]
We are committed to making our proposed model and methods publicly available following the completion of our thesis and ensuring its contribution to the research community. Additionally, we are dedicated to accurately citing your contribute and acknowledging your assistance during our presentations.
Sincerely,
Nguyen Xuan Tuan
@hysts Excuse me Sir, I really need your help
Hi
@tuansunday05
This Space seems to run at a reasonable speed even on cpu-basic, and I wondered if it really needs a GPU, so I've asked other HF members to review this as well.
That being said, if it can run on ZeroGPU, maybe we can assign ZeroGPU as a grant. (I'm not sure if this Space is compatible with Zero, though.)
I just assigned ZeroGPU to this Space. Can you check the org card of the ZeroGPU explorers org to see if your Space can run on Zero.
@hysts Thank you so much for your assistance <3, my space can run very well on this ZeroGPU that your team provied .