cheeun/cadyq

[ECCV2022] Official Code for the "CADyQ: Content-Aware Dynamic Quantization for Image Super Resolution"

cadyq's Issues

Hardware support for inference

Hi, it is not mentioned in the paper that how to implement the NVIDIA support for the quantized model to inference. Could you please give some explanations? Thks very much!

New Super-Resolution Benchmarks

Hello,

MSU Graphics & Media Lab Video Group has recently launched two new Super-Resolution Benchmarks.

• Video Upscalers Benchmark: Quality Enhancement determines the best upscaling methods for increasing video resolution and improving visual quality.
• Super-Resolution for Video Compression benchmark aims to test Super-Resolution methods on compressed videos and select the best model for each video codec standard.

If you are interested in participating, you can add your algorithm following the submission steps:

• Submit for Video Upscalers Benchmark: Quality Enhancement
• Submit for Super-Resolution for Video Compression benchmark

We would be grateful for your feedback on our work!

It's a nice work to SR. But I face this problem. For LR_bicubic, the name of pictures have the format of img001x4.png.

Questions about FQR

Hi, i have some questions about FQR!

Are avg_bit and FQR the same thing?
Does FQR only consider the residual blocks?
Aren't you considering the 1x1 conv inside the CARN?

Could you please give some explanations?
Thank you very much

Wonderful work！But its markdown doesn't contain a file named 'div2k_valid'. Can you add a Google Link To It？Thanks

Questions about BitOPs

Hi, I have some questions about BitOPs!

The paper says BitOPs is measured by generating a 720×1820 picture，so is it Urban100 datasets or others？
I implemented the bitops code myself，The results were the same except for cadyq's. The algorithm for bitops is not in the code，How do you implement BitOPs？ Based on thop or other open source libraries？

Could you please give some explanations?
Thank you very much!

I want to inference my own image .

Now I want to inference my image in your model . I want to trace your model when I set args.fully is True , there is AttributeError: module 'model.carn_pams' has no attribute 'PAMS_BasicBlock' . But when I run test.py it's ok .

which is correct checkpoint for inference ?

It's a nice work for SR ! But when I run test_cadyq_image.sh and student_weights for your checkpoint , all have missing key error .

Error(s) in loading state_dict for CARN_CADyQ:

It's a nice work for SR ! I have met the issue that RuntimeError: Error(s) in loading state_dict for EDSR_CADyQ:
Missing key(s) in state_dict: "body.9.weight", "body.9.bias".
Unexpected key(s) in state_dict: "body.10.bitsel1.quant_bit1.alpha",......when i use the command of test after i trained the model of EDSR for cadyq.
How can i solve this problem?