Visualizing the Process of Hieroglyphic Evolution with Inter-Frame Attention

This study constructed a data set named Oracle to Simplified Chinese (OtSC105) for the evolution of hieroglyphs. Using an inter-frame attention extraction algorithm, images were generated to achieve the visualization of hieroglyphic evolution from oracle bone inscriptions to modern Chinese characters.

Getting Started

We need the following listed environments：

• torch 1.8.0
• python 3.8
• skimage 0.19.2
• numpy 1.23.1
• opencv-python 4.6.0
• timm 0.6.11
• tqdm

Play with Demos

1. Download the model checkpoints and put the ckpt folder into the root dir.
2. Download the dataset OtSC105 or your own pictures, then put it into the folder 'Sources'.
3. Run the following commands to generate Nx (arbitrary) frame interpolation demos:

python Generate_from_dataset.py --model /ours_t/ours --InputPath /Your/Dataset/Path --OutputPath /Your/Output/Path --n /Insert/Frames

Evaluation

1. Using the testbench in the folder of 'Sources'
2. Download the model checkpoints and put the ckpt folder into the root dir.
3. For 2x interpolation benchmarks:

   python OtSC105.py --model /ours/ours_small --path /Your/Dataset/Path