init

LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2021 Alex
+Copyright (c) 2021 Alexander Spirin
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

faces.py

@@ -0,0 +1,82 @@
+"""
+    Code related to face detection and manipulation
+"""
+
+#pip install facenet_pytorch
+
+from facenet_pytorch import MTCNN
+mtcnn = MTCNN(image_size=256, margin=80)
+
+# simplest ye olde trustworthy MTCNN for face detection with landmarks
+def detect(img):
+
+        # Detect faces
+        batch_boxes, batch_probs, batch_points = mtcnn.detect(img, landmarks=True)
+        # Select faces
+        if not mtcnn.keep_all:
+            batch_boxes, batch_probs, batch_points = mtcnn.select_boxes(
+                batch_boxes, batch_probs, batch_points, img, method=mtcnn.selection_method
+            )
+
+        return batch_boxes, batch_points
+
+# my version of isOdd, should make a separate repo for it :D
+def makeEven(_x):
+  return _x if (_x % 2 == 0) else _x+1
+
+# the actual scaler function
+def scale(boxes, _img, max_res=1_500_000, target_face=256, fixed_ratio=0, max_upscale=2, VERBOSE=False):
+
+    x, y = _img.size
+
+    ratio = 2 #initial ratio
+
+    #scale to desired face size
+    if (boxes is not None):
+      if len(boxes)>0:
+        ratio = target_face/max(boxes[0][2:]-boxes[0][:2]); 
+        ratio = min(ratio, max_upscale)
+        if VERBOSE: print('up by', ratio)
+
+    if fixed_ratio>0:
+      if VERBOSE: print('fixed ratio')
+      ratio = fixed_ratio
+
+    x*=ratio
+    y*=ratio
+
+    #downscale to fit into max res 
+    res = x*y
+    if res > max_res:
+      ratio = pow(res/max_res,1/2); 
+      if VERBOSE: print(ratio)
+      x=int(x/ratio)
+      y=int(y/ratio)
+
+    #make dimensions even, because usually NNs fail on uneven dimensions due skip connection size mismatch
+    x = makeEven(int(x))
+    y = makeEven(int(y))
+
+    size = (x, y)
+
+    return _img.resize(size)
+
+""" 
+    A useful scaler algorithm, based on face detection.
+    Takes PIL.Image, returns a uniformly scaled PIL.Image
+
+    boxes: a list of detected bboxes
+    _img: PIL.Image
+    max_res: maximum pixel area to fit into. Use to stay below the VRAM limits of your GPU.
+    target_face: desired face size. Upscale or downscale the whole image to fit the detected face into that dimension.
+    fixed_ratio: fixed scale. Ignores the face size, but doesn't ignore the max_res limit.
+    max_upscale: maximum upscale ratio. Prevents from scaling images with tiny faces to a blurry mess.
+"""
+
+def scale_by_face_size(_img, max_res=1_500_000, target_face=256, fix_ratio=0, max_upscale=2, VERBOSE=False):
+    boxes = None
+    boxes, _ = detect(_img)
+    if VERBOSE: print('boxes',boxes)
+    img_resized = scale(boxes, _img, max_res, target_face, fix_ratio, max_upscale, VERBOSE)
+    return img_resized
+

ffmpeg.py

@@ -0,0 +1,32 @@
+"""
+    Combining frames into a video, then adding sound from the source video.
+
+    frames: a glob-like pattern, for example '/content/out_resized/frame*.jpg'
+    vid_name: path to the input video
+    fps: fps :D
+    out_fname: output video name. Use mp4 container by default.
+"""
+
+ffmpeg -f image2 -r {fps} -pattern_type glob -i "{frames}"  -i "{vid_name}" -c:v copy -c:a aac -map 0:v:0 \
+ -map 1:a:0 -shortest  -vcodec libx264 -r {fps} -crf 18  -pix_fmt yuv420p   "{out_fname}"
+
+"""
+    The simplest way to combine frames into video
+
+    frames: a glob-like pattern, for example '/content/out_resized/frame*.jpg'
+    fps: fps :D
+    out_fname: output video name. Use mp4 container by default.
+"""
+
+ffmpeg -pattern_type glob -i "{frames}" -filter:v fps={fps} "{out_fname}"
+
+"""
+    The simplest way to split a video into frames
+
+    vid_name: input video'
+    out_dir: output folder. 
+
+    You can change the pattern if you need more digits (increase the number in %05d pattern), or change the filename. 
+"""
+
+ffmpeg -v quiet -i "{vid_name}" "{out_dir}/frame_%05d.jpg"

google_colab.py

@@ -0,0 +1,30 @@
+"""
+    Some of the colab snippets I use on a daily basis.
+    I've decided to store those as .py for faster access from github and easier imports.
+"""
+"""
+    Zips a folder and downloads it. Can also copy the archive to a google drive path (or any other available to your colab instance)
+    Use to automatically download results after some afk training or inference, before the instance resets.
+
+    path: path to a folder to download. Can contain wildcards
+    prefix: a name of your new archive
+    drive: a folder to copy the archive to
+"""
+from google.colab import files
+import time
+
+def download(path, prefix, drive=''):
+  print('\nCкачиваем...\n')
+  timestamp = round(time.time())
+  !zip -r /content/{prefix}-{timestamp}.zip {path}
+  files.download(f"/content/{prefix}-{timestamp}.zip")
+  if drive!='':
+      !cp "/content/{prefix}-{timestamp}.zip" {drive}
+
+
+"""
+    It beeps. Use to give yourself a signal.
+"""
+from google.colab import output
+def beep():
+  output.eval_js('new Audio("https://upload.wikimedia.org/wikipedia/commons/0/05/Beep-09.ogg").play()')