-
Alot of these networks are still too heavy for cpu inference would you be open to a PR to add these models ?
-
![image](https://github.com/user-attachments/assets/f2dc67b1-9762-4d08-a0e5-fc10ed0d937c)
When using face_upsample, this situation may occur if the face is blocked by other objects.
-
[VAIQ_ERROR][QUANTIZER_TORCH_SHAPE_MISMATCH]: output shape of FSRCNN::FSRCNN/ConvTranspose2d[last_part]/ret([1, 80, 80, 1]) is different from the output shape of XIR ([1, 79, 79, 1]).
Why is there a …
-
I just run 'python val.py --config_path ./configs/basicvsr/val.yaml --save_metrics ./output/quality_metrics.json'.
But i got :
raceback (most recent call last):
File "val.py", line 51, in
…
-
Hello!
I even succeeded in qauntizing FSRCNN models.
I'm using U200 models
However I don't know why there was a program error.
```
(vitis-ai-tensorflow2) Vitis-AI /workspace/juwon/FSRCNN-tf2 >…
-
hello, i have an issue is about the compared model of FSRCNN(origin), and your model of FSRCNN training by PISR.
the origin model of fsrcnn is trained without div2k, but your model has.
so, maybe y…
-
hello,do you have the FSRCNN's Train[caffe+Matlab] code provided by the author?
Can I have a copy?
Thanks!
-
你好,感谢您提供相关代码。
在本仓库运行时,使用您提供的预训练模型,测试butterfly 2/3倍scale时,CPU耗时700ms,GPU(NVIDIA-SMI 515.65.01 Driver Version: 515.65.01 CUDA Version: 11.7)耗时400ms,
与论文中的结论27fps有差异,是什么原因呢
-
### Expected behavior 预期的功能
There is an additional set of FSRCNNX shaders named as [FSRCNNX-Distort](https://github.com/HelpSeeker/FSRCNN-TensorFlow/releases/tag/1.1_distort) similar to LineArt but a…
-
Hi! I meet some problems.
**WARNING: [HLS 200-885] Unable to schedule 'load' operation ('__Val2__') on array 'buffer_data_V' due to limited memory ports. Please consider using a memory core with mo…