Spaces:
Running
Running
File size: 8,936 Bytes
9e1ec94 7fbe9b9 520449a 9e1ec94 fae5be7 61c9bd4 a3296a3 9e1ec94 2b7c418 9e1ec94 ece198d cdddf52 9e1ec94 40a4317 9e1ec94 b2d74ab 9e1ec94 aad61cc cc08c34 b2d74ab 61c9bd4 520449a 7fbe9b9 520449a 7fbe9b9 61c9bd4 7fbe9b9 61c9bd4 7fbe9b9 61c9bd4 b2d74ab 55bf202 9e1ec94 55bf202 9e1ec94 fae5be7 9e1ec94 55bf202 9e1ec94 55bf202 9e1ec94 fae5be7 a3296a3 7fbe9b9 a3296a3 7fbe9b9 a3296a3 b2d74ab 9e1ec94 22320ed 9e1ec94 b2d74ab 61c9bd4 b2d74ab a3296a3 7d2266d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 |
import os
import io
import time
import json
import math
import base64
from uuid import uuid4
from PIL import Image as PILImage
os.environ["DEEPFACE_HOME"] = "."
import pyzipper
import numpy as np
import gradio as gr
from annoy import AnnoyIndex
from deepface import DeepFace
index = AnnoyIndex(512, "euclidean")
index.load(f"face.db")
ANNOY_INDEX = json.load(open(f"face.json"))
with pyzipper.AESZipFile('persons.zip') as zf:
password = os.getenv("VISAGE_KEY","").encode('ascii')
zf.setpassword(password)
PERFORMER_DB = json.loads(zf.read('performers.json'))
## Prediction functions
def image_search_performer(image, threshold=20.0, results=3):
"""Search for a performer in an image
Returns a list of performers with at least following keys:
- id: the performer's id
- distance: the distance between the face in the image and the performer's face
- confidence: a confidence score between 0 and 100
- hits: the number of times the performer was found in our database
"""
image_array = np.array(image)
face = DeepFace.represent(img_path = image_array, detector_backend='retinaface', model_name='Facenet512', normalization="Facenet2018")[0]['embedding']
return search_performer(face, threshold, results)
def image_search_performers(image, threshold=20.0, results=3):
image_array = np.array(image)
response = []
t = time.time()
try:
faces = DeepFace.represent(img_path = image_array, detector_backend='retinaface', model_name='Facenet512', normalization="Facenet2018")
# faces = DeepFace.represent(img_path = image_array, detector_backend='yolov8', model_name='Facenet512', normalization="Facenet2018")
# faces = DeepFace.represent(img_path = image_array, detector_backend='mtcnn', model_name='Facenet512', normalization="Facenet2018")
except ValueError as e:
print(e)
raise gr.Error("No faces found in the image")
print(f"Time to find faces: {time.time() - t}")
for face in faces:
embedding = face['embedding']
area = face['facial_area']
confidence = face['face_confidence']
cimage = image.crop((area['x'], area['y'], area['x'] + area['w'], area['y'] + area['h']))
buf = io.BytesIO()
cimage.save(buf, format='JPEG')
im_b64 = base64.b64encode(buf.getvalue()).decode('ascii')
response.append({
'image': im_b64,
'confidence': confidence,
'performers': search_performer(embedding, threshold, results)
})
return response
def vector_search_performer(vector_json, threshold=20.0, results=3):
"""Search for a performer from a vector
The vector should be created with Deepface and should be a 512 vector.
For best results use the following settings:
- detector_backend: retinaface
- model: Facenet512
- normalization: Facenet2018
Returns a list of performers with at least following keys:
- id: the performer's id
- distance: the distance between the face in the image and the performer's face
- confidence: a confidence score between 0 and 100
- hits: the number of times the performer was found in our database
"""
vector = np.array(json.loads(vector_json))
return search_performer(vector, threshold, results)
def search_performer(vector, threshold=20.0, results=3):
threshold = threshold or 20.0
results = results or 3
ids, distances = index.get_nns_by_vector(
vector, 50, search_k=100000, include_distances=True
)
persons = {}
for p, distance in zip(ids, distances):
id = ANNOY_INDEX[p]
if id in persons:
persons[id]["hits"] += 1
persons[id]["distance"] -= 0.5
persons[id]["confidence"] = normalize_confidence_from_distance(persons[id]["distance"], threshold)
continue
persons[id] = {
"id": id,
"distance": round(distance, 2),
"confidence": normalize_confidence_from_distance(distance, threshold),
"hits": 1,
}
if id in PERFORMER_DB:
persons[id].update(PERFORMER_DB.get(id))
persons = sorted(persons.values(), key=lambda x: x["distance"])
# persons = [p for p in persons if p["distance"] < threshold]
return persons[:results]
def normalize_confidence_from_distance(distance, threshold=20.0):
"""Normalize confidence to 0-100 scale"""
confidence = face_distance_to_conf(distance, threshold)
return int(((confidence - 0.0) / (1.0 - 0.0)) * (100.0 - 0.0) + 0.0)
def face_distance_to_conf(face_distance, face_match_threshold=20.0):
"""Using a face distance, calculate a similarity confidence value"""
if face_distance > face_match_threshold:
# The face is far away, so give it a low confidence
range = (1.0 - face_match_threshold)
linear_val = (1.0 - face_distance) / (range * 2.0)
return linear_val
else:
# The face is close, so give it a high confidence
range = face_match_threshold
linear_val = 1.0 - (face_distance / (range * 2.0))
# But adjust this value by a curve so that we don't get a linear
# transition from close to far. We want it to be more confident
# the closer it is.
return linear_val + ((1.0 - linear_val) * math.pow((linear_val - 0.5) * 2, 0.2))
def find_faces_in_sprite(image, vtt):
vtt = base64.b64decode(vtt.replace("data:text/vtt;base64,", ""))
sprite = PILImage.fromarray(image)
results = []
for i, (left, top, right, bottom, time_seconds) in enumerate(getVTToffsets(vtt)):
cut_frame = sprite.crop((left, top, left + right, top + bottom))
faces = DeepFace.extract_faces(np.asarray(cut_frame), detector_backend="mediapipe", enforce_detection=False, align=False)
faces = [face for face in faces if face['confidence'] > 0.6]
if faces:
size = faces[0]['facial_area']['w'] * faces[0]['facial_area']['h']
data = {'id': str(uuid4()), "offset": (left, top, right, bottom), "frame": i, "time": time_seconds, 'size': size}
results.append(data)
# sort by size
# results = sorted(results, key=lambda x: x['size'], reverse=True)
return results
def getVTToffsets(vtt):
time_seconds = 0
left = top = right = bottom = None
for line in vtt.decode("utf-8").split("\n"):
line = line.strip()
if "-->" in line:
# grab the start time
# 00:00:00.000 --> 00:00:41.000
start = line.split("-->")[0].strip().split(":")
# convert to seconds
time_seconds = (
int(start[0]) * 3600
+ int(start[1]) * 60
+ float(start[2])
)
left = top = right = bottom = None
elif "xywh=" in line:
left, top, right, bottom = line.split("xywh=")[-1].split(",")
left, top, right, bottom = (
int(left),
int(top),
int(right),
int(bottom),
)
else:
continue
if not left:
continue
yield left, top, right, bottom, time_seconds
image_search = gr.Interface(
fn=image_search_performer,
inputs=[
gr.components.Image(),
gr.components.Slider(label="threshold",minimum=0.0, maximum=30.0, value=20.0),
gr.components.Slider(label="results", minimum=0, maximum=50, value=3, step=1),
],
outputs=gr.outputs.JSON(label=""),
title="Who is in the photo?",
description="Upload an image of a person and we'll tell you who it is.",
)
image_search_multiple = gr.Interface(
fn=image_search_performers,
inputs=[
gr.components.Image(type='pil'),
gr.components.Slider(label="threshold",minimum=0.0, maximum=30.0, value=20.0),
gr.components.Slider(label="results", minimum=0, maximum=50, value=3, step=1),
],
outputs=gr.outputs.JSON(label=""),
title="Who is in the photo?",
description="Upload an image of a person(s) and we'll tell you who it is.",
)
vector_search = gr.Interface(
fn=vector_search_performer,
inputs=[
gr.components.Textbox(),
gr.components.Slider(label="threshold",minimum=0.0, maximum=30.0, value=20.0),
gr.components.Slider(label="results", minimum=0, maximum=50, value=3, step=1),
],
outputs=gr.outputs.JSON(label=""),
title="Who is in the photo?",
description="512 vector created with deepface of a person and we'll tell you who it is.",
)
faces_in_sprite = gr.Interface(
fn=find_faces_in_sprite,
inputs=[
gr.Image(),
gr.Textbox(label="VTT file")
],
outputs=gr.JSON(label=""),
)
gr.TabbedInterface([image_search, image_search_multiple, vector_search, faces_in_sprite]).queue().launch(enable_queue=True, server_name="0.0.0.0")
|