app.py

import os
import pickle
import sys

import subprocess
import tarfile

import subprocess

if os.environ.get('SYSTEM') == 'spaces':
    subprocess.call('pip install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu113'.split())
    
    
import imageio
import numpy as np
import scipy.interpolate
import torch
from tqdm import tqdm
import gradio as gr 
from huggingface_hub import hf_hub_download


def layout_grid(img, grid_w=None, grid_h=1, float_to_uint8=True, chw_to_hwc=True, to_numpy=True):
    batch_size, channels, img_h, img_w = img.shape
    if grid_w is None:
        grid_w = batch_size // grid_h
    assert batch_size == grid_w * grid_h
    if float_to_uint8:
        img = (img * 127.5 + 128).clamp(0, 255).to(torch.uint8)
    img = img.reshape(grid_h, grid_w, channels, img_h, img_w)
    img = img.permute(2, 0, 3, 1, 4)
    img = img.reshape(channels, grid_h * img_h, grid_w * img_w)
    if chw_to_hwc:
        img = img.permute(1, 2, 0)
    if to_numpy:
        img = img.cpu().numpy()
    return img




network_pkl='braingan-400.pkl'
with open(network_pkl, 'rb') as f:
    G = pickle.load(f)['G_ema'] 

def predict(Seed,choices):
  device = torch.device('cuda')
  G.eval()
  G.to(device)
  shuffle_seed=None
  w_frames=60*4
  kind='cubic' 
  num_keyframes=None
  wraps=2
  psi=1 
  device=torch.device('cuda')

  
  if choices=='4x2':
    grid_w = 4
    grid_h = 2
    s1=Seed
    seeds=(np.arange(s1-16,s1)).tolist()
  if choices=='2x1':
    grid_w = 2
    grid_h = 1
    s1=Seed
    seeds=(np.arange(s1-4,s1)).tolist()


  mp4='ex.mp4'
  truncation_psi=1
  num_keyframes=None


  if num_keyframes is None:
      if len(seeds) % (grid_w*grid_h) != 0:
          raise ValueError('Number of input seeds must be divisible by grid W*H')
      num_keyframes = len(seeds) // (grid_w*grid_h)

  all_seeds = np.zeros(num_keyframes*grid_h*grid_w, dtype=np.int64)
  for idx in range(num_keyframes*grid_h*grid_w):
      all_seeds[idx] = seeds[idx % len(seeds)]

  if shuffle_seed is not None:
      rng = np.random.RandomState(seed=shuffle_seed)
      rng.shuffle(all_seeds)

  zs = torch.from_numpy(np.stack([np.random.RandomState(seed).randn(G.z_dim) for seed in all_seeds])).to(device)
  ws = G.mapping(z=zs, c=None, truncation_psi=psi)
  _ = G.synthesis(ws[:1]) # warm up
  ws = ws.reshape(grid_h, grid_w, num_keyframes, *ws.shape[1:])

  # Interpolation.
  grid = []
  for yi in range(grid_h):
      row = []
      for xi in range(grid_w):
          x = np.arange(-num_keyframes * wraps, num_keyframes * (wraps + 1))
          y = np.tile(ws[yi][xi].cpu().numpy(), [wraps * 2 + 1, 1, 1])
          interp = scipy.interpolate.interp1d(x, y, kind=kind, axis=0)
          row.append(interp)
      grid.append(row)

  # Render video.
  video_out = imageio.get_writer(mp4, mode='I', fps=60, codec='libx264')
  for frame_idx in tqdm(range(num_keyframes * w_frames)):
      imgs = []
      for yi in range(grid_h):
          for xi in range(grid_w):
              interp = grid[yi][xi]
              w = torch.from_numpy(interp(frame_idx / w_frames)).to(device)
              img = G.synthesis(ws=w.unsqueeze(0), noise_mode='const')[0]
              imgs.append(img)
      video_out.append_data(layout_grid(torch.stack(imgs), grid_w=grid_w, grid_h=grid_h))
  video_out.close()
  return 'ex.mp4'



choices=['4x2','2x1']
interface=gr.Interface(fn=predict, title="Brain MR Image Generation with StyleGAN-2",
                       description = "",
                       article = "Author: S.Serdar Helli",
                       inputs=[gr.inputs.Slider( minimum=16, maximum=2**10,label='Seed'),gr.inputs.Radio( choices=choices,  default='4x2',label='Image Grid')],
                       outputs=gr.outputs.Video(label='Video'))


interface.launch(debug=True)