app.py

import json
import operator
import gradio as gr
import torchvision
from typing import Tuple, Dict
from facetorch import FaceAnalyzer
from facetorch.datastruct import ImageData
from omegaconf import OmegaConf
from torch.nn.functional import cosine_similarity


cfg = OmegaConf.load("config.merged.yml")
analyzer = FaceAnalyzer(cfg.analyzer)

def get_sim_dict_str(response: ImageData, pred_name: str = "verify", index: int = 0)-> str:
    base_emb = response.faces[index].preds[pred_name].logits
    sim_dict = {face.indx: cosine_similarity(base_emb, face.preds[pred_name].logits, dim=0).item() for face in response.faces}
    sim_dict_sort = dict(sorted(sim_dict.items(), key=operator.itemgetter(1),reverse=True))
    sim_dict_sort_str = str(sim_dict_sort)
    return sim_dict_sort_str


def inference(path_image: str) -> Tuple:
    response = analyzer.run(
        path_image=path_image,
        batch_size=cfg.batch_size,
        fix_img_size=cfg.fix_img_size,
        return_img_data=cfg.return_img_data,
        include_tensors=cfg.include_tensors,
        path_output=None,
    )
    
    pil_image = torchvision.transforms.functional.to_pil_image(response.img)
    
    fer_dict_str = str({face.indx: face.preds["fer"].label for face in response.faces})
    deepfake_dict_str = str({face.indx: face.preds["deepfake"].label for face in response.faces})
    response_str = str(response)
    
    sim_dict_str_embed = get_sim_dict_str(response, pred_name="embed", index=0)
    sim_dict_str_verify = get_sim_dict_str(response, pred_name="verify", index=0)
    
    out_tuple = (pil_image, fer_dict_str, deepfake_dict_str, sim_dict_str_embed, sim_dict_str_verify, response_str)
    return out_tuple


title = "facetorch-app"
description = "Demo of facetorch, a Python library that uses pre-trained deep neural networks for face detection, representation learning, verification, expression recognition, deepfake detection, and 3D alignment. Try selecting one of the example images or upload your own. This work would not be possible without the original work of the researchers and engineers who trained the models (sources and credits can be found in the facetorch repository). Use responsibly."
article = "<p style='text-align: center'><a href='https://github.com/tomas-gajarsky/facetorch' style='text-align:center' target='_blank'>facetorch GitHub repository</a><a href='https://pypi.org/project/facetorch/'><img src='https://img.shields.io/pypi/v/facetorch' style='text-align:center' alt='PyPI'></a><a href='https://anaconda.org/conda-forge/facetorch' style='text-align:center'><img src='https://img.shields.io/conda/vn/conda-forge/facetorch' alt='Conda (channel only)'></a><a href='https://raw.githubusercontent.com/tomas-gajarsky/facetorch/main/LICENSE' style='text-align:center'><img src='https://img.shields.io/pypi/l/facetorch' alt='PyPI - License' class='center'></a><img src='https://raw.githubusercontent.com/tomas-gajarsky/facetorch/main/data/facetorch-logo-64.png' alt='' class='center' title='facetorch logo'></p>"

demo=gr.Interface(
    inference,
    [gr.inputs.Image(label="Input", type="filepath")],
    [gr.outputs.Image(type="pil", label="Face Detection and 3D Landmarks"),
     gr.outputs.Textbox(label="Facial Expression Recognition"),
     gr.outputs.Textbox(label="DeepFake Detection"),
     gr.outputs.Textbox(label="Cosine similarity of Face Representation Embeddings"),
     gr.outputs.Textbox(label="Cosine similarity of Face Verification Embeddings"),
     gr.outputs.Textbox(label="Response")],
    title=title,
    description=description,
    article=article,
    examples=[["./test5.jpg"], ["./test.jpg"], ["./test4.jpg"], ["./test2.jpg"], ["./test8.jpg"], ["./test6.jpg"], ["./test3.jpg"]],
)
demo.launch(server_name="0.0.0.0", server_port=7860, debug=True)