videomae-vis / app.py
SauravMaheshkar's picture
feat: initial commit
65947b1 unverified
raw
history blame
1.41 kB
import gradio as gr
import torch
from augmentations import get_videomae_transform
from models import load_model
from utils import create_plot, get_frames, get_videomae_outputs, prepare_frames_masks
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
transform = get_videomae_transform()
def get_visualisations(mask_ratio, video_path):
frames, ids = get_frames(path=video_path, transform=transform)
model, masks, patch_size = load_model(
path="assets/checkpoint.pth",
mask_ratio=mask_ratio,
device=device,
)
with torch.no_grad():
frames, masks = prepare_frames_masks(frames, masks, device)
outputs = model(frames, masks)
visualisations = get_videomae_outputs(
frames=frames,
masks=masks,
outputs=outputs,
ids=ids,
patch_size=patch_size,
device=device,
)
return create_plot(visualisations)
with gr.Blocks() as app:
video = gr.Video(
value="assets/example.mp4",
)
mask_ratio_slider = gr.Slider(
minimum=0.25, maximum=0.95, step=0.05, value=0.75, label="masking ratio"
)
btn = gr.Button("Run")
btn.click(
get_visualisations,
inputs=[mask_ratio_slider, video],
outputs=gr.Plot(label="VideoMAE Outputs", format="png"),
)
if __name__ == "__main__":
app.launch()