Instruction for calculating metrics

Prepare the frames

For fast evaluation, we measure 1 frames every 10 frames, that is, the 0-th frame, 10-th frame, 20-th frame, etc. will participate the metrics calculation. This can be achieved by sample_interval argument.

python scripts/inference_animesr_frames.py -i AVC-RealLQ-ROOT -n AnimeSR_v1-PaperModel --expname animesr_v1_si10 --sample_interval 10

MANIQA calculation

requirements

pip install timm==0.5.4

checkpoint

we use the ensemble model provided by the authors to compute MANIQA. You can download the checkpoint from Ondrive.

inference:

# cd into scripts/metrics/MANIQA
python inference_MANIQA.py --model_path MANIQA_CKPT_YOU_JUST_DOWNLOADED --input_dir ../../../results/animesr_v1_si10/frames --output_dir output/ensemble_attentionIQA2_finetune_e2/AnimeSR_v1_si10

note that the result has certain randomness, but the error should be relatively small.

license

the MANIQA codes&checkpoint are original from MANIQA and @TianheWu.