Upload folder using huggingface_hub

.gitattributes

@@ -25,7 +25,6 @@
 *.safetensors filter=lfs diff=lfs merge=lfs -text
 saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
 *.tflite filter=lfs diff=lfs merge=lfs -text
 *.tgz filter=lfs diff=lfs merge=lfs -text
 *.wasm filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,2 @@
+__pycache__/
+venv/

README.md

@@ -1,13 +1,12 @@
 ---
 title: Image Similarity
-emoji: 🏃
-colorFrom: gray
-colorTo: blue
+emoji: 🐨
+colorFrom: red
+colorTo: green
 sdk: gradio
-sdk_version: 3.45.1
+sdk_version: 3.16.0
 app_file: app.py
 pinned: false
-license: mit
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,32 @@
+import gradio as gr
+import os
+import random
+from src.model import simlarity_model as model
+from src.similarity.similarity import Similarity
+
+similarity = Similarity()
+models = similarity.get_models()
+
+def check(img_main, img_1, img_2, model_idx):    
+    result = similarity.check_similarity([img_main, img_1, img_2], models[model_idx])
+    return result
+
+with gr.Blocks() as demo:
+    gr.Markdown('Checking Image Similarity')
+    img_main = gr.Text(label='Main Image', placeholder='https://myimage.jpg')
+    
+    gr.Markdown('Images to check')
+    img_1 = gr.Text(label='1st Image', placeholder='https://myimage_1.jpg')
+    img_2 = gr.Text(label='2nd Image', placeholder='https://myimage_2.jpg')
+
+    gr.Markdown('Choose the model')
+    model = gr.Dropdown([m.name for m in models], label='Model', type='index')
+    
+    gallery = gr.Gallery(
+            label="Generated images", show_label=False, elem_id="gallery"
+        ).style(grid=[2], height="auto")
+
+    submit_btn = gr.Button('Check Similarity')
+    submit_btn.click(fn=check,inputs=[img_main, img_1, img_2, model], outputs=gallery)
+
+demo.launch()    

requirements.txt

@@ -0,0 +1,103 @@
+absl-py==1.3.0
+aiohttp==3.8.3
+aiosignal==1.3.1
+altair==4.2.0
+anyio==3.6.2
+astunparse==1.6.3
+async-timeout==4.0.2
+attrs==22.2.0
+cachetools==5.2.0
+certifi==2022.12.7
+charset-normalizer==2.1.1
+click==8.1.3
+contourpy==1.0.6
+cycler==0.11.0
+entrypoints==0.4
+fastapi==0.88.0
+ffmpy==0.3.0
+filelock==3.9.0
+flatbuffers==23.1.4
+fonttools==4.38.0
+frozenlist==1.3.3
+fsspec==2022.11.0
+gast==0.4.0
+google-auth==2.15.0
+google-auth-oauthlib==0.4.6
+google-pasta==0.2.0
+gradio==3.16.0
+grpcio==1.51.1
+h11==0.14.0
+h5py==3.7.0
+httpcore==0.16.3
+httpx==0.23.3
+huggingface-hub==0.11.1
+idna==3.4
+importlib-metadata==6.0.0
+importlib-resources==5.10.2
+Jinja2==3.1.2
+jsonschema==4.17.3
+keras==2.11.0
+kiwisolver==1.4.4
+libclang==14.0.6
+linkify-it-py==1.0.3
+Markdown==3.4.1
+markdown-it-py==2.1.0
+MarkupSafe==2.1.1
+matplotlib==3.6.2
+mdit-py-plugins==0.3.3
+mdurl==0.1.2
+multidict==6.0.4
+numpy==1.24.1
+nvidia-cublas-cu11==11.10.3.66
+nvidia-cuda-nvrtc-cu11==11.7.99
+nvidia-cuda-runtime-cu11==11.7.99
+nvidia-cudnn-cu11==8.5.0.96
+oauthlib==3.2.2
+opt-einsum==3.3.0
+orjson==3.8.4
+packaging==22.0
+pandas==1.5.2
+Pillow==9.4.0
+pkgutil_resolve_name==1.3.10
+protobuf==3.19.6
+pyasn1==0.4.8
+pyasn1-modules==0.2.8
+pycryptodome==3.16.0
+pydantic==1.10.4
+pydub==0.25.1
+pyparsing==3.0.9
+pyrsistent==0.19.3
+python-dateutil==2.8.2
+python-multipart==0.0.5
+pytz==2022.7
+PyYAML==6.0
+regex==2022.10.31
+requests==2.28.1
+requests-oauthlib==1.3.1
+rfc3986==1.5.0
+rsa==4.9
+six==1.16.0
+sniffio==1.3.0
+starlette==0.22.0
+tensorboard==2.11.0
+tensorboard-data-server==0.6.1
+tensorboard-plugin-wit==1.8.1
+tensorflow==2.11.0
+tensorflow-estimator==2.11.0
+tensorflow-hub==0.12.0
+tensorflow-io-gcs-filesystem==0.29.0
+termcolor==2.2.0
+tokenizers==0.13.2
+toolz==0.12.0
+torch==1.13.1
+tqdm==4.64.1
+transformers==4.25.1
+typing_extensions==4.4.0
+uc-micro-py==1.0.1
+urllib3==1.26.13
+uvicorn==0.20.0
+websockets==10.4
+Werkzeug==2.2.2
+wrapt==1.14.1
+yarl==1.8.2
+zipp==3.11.0

src/model/similarity_interface.py

@@ -0,0 +1,3 @@
+class SimilarityInterface:
+    def extract_feature(img):
+        return []

src/model/simlarity_model.py

@@ -0,0 +1,9 @@
+from dataclasses import dataclass
+from .similarity_interface import SimilarityInterface
+
+@dataclass
+class SimilarityModel:
+    name: str 
+    image_size: int
+    model_cls: SimilarityInterface
+    image_input_type: str = 'array'

src/similarity/model_implements/bit.py

@@ -0,0 +1,13 @@
+import tensorflow_hub as hub
+import numpy as np 
+
+class BigTransfer:
+
+    def __init__(self):
+        self.module = hub.KerasLayer("https://tfhub.dev/google/bit/m-r50x1/1")
+
+    def extract_feature(self, imgs):
+        features = []
+        for img in imgs:
+            features.append(np.squeeze(self.module(img)))
+        return features

src/similarity/model_implements/mobilenet_v3.py

@@ -0,0 +1,14 @@
+import tensorflow_hub as hub
+import numpy as np 
+
+class ModelnetV3():
+    def __init__(self):
+        module_handle = "https://tfhub.dev/google/imagenet/mobilenet_v3_large_100_224/feature_vector/5" 
+        self.module = hub.load(module_handle)
+
+    def extract_feature(self, imgs):
+        print('getting with ModelnetV3...')
+        features = []
+        for img in imgs:
+            features.append(np.squeeze(self.module(img)))
+        return features

src/similarity/model_implements/vit_base.py

@@ -0,0 +1,20 @@
+from transformers import ViTFeatureExtractor, ViTModel
+from PIL import Image
+import numpy as np 
+import torch
+
+class VitBase():
+
+    def __init__(self):
+        self.feature_extractor = ViTFeatureExtractor.from_pretrained('google/vit-base-patch16-224-in21k')
+        self.model = ViTModel.from_pretrained('google/vit-base-patch16-224-in21k')
+    
+    def extract_feature(self, imgs):
+        features = []
+        for img in imgs:
+            inputs = self.feature_extractor(images=img, return_tensors="pt")
+            with torch.no_grad():
+                outputs = self.model(**inputs)
+            last_hidden_states =  outputs.last_hidden_state            
+            features.append(np.squeeze(last_hidden_states.numpy()).flatten())            
+        return features

src/similarity/similarity.py

@@ -0,0 +1,35 @@
+from src.model import simlarity_model as model 
+from src.util import image as image_util
+from src.util import matrix
+from .model_implements.mobilenet_v3 import ModelnetV3
+from .model_implements.vit_base import VitBase
+from .model_implements.bit import BigTransfer
+
+
+class Similarity:
+    def get_models(self):
+        return [
+            model.SimilarityModel(name= 'Mobilenet V3', image_size= 224, model_cls = ModelnetV3()),
+            model.SimilarityModel(name= 'Big Transfer (BiT)', image_size= 224, model_cls = BigTransfer()),
+            model.SimilarityModel(name= 'Vision Transformer', image_size= 224, model_cls = VitBase(), image_input_type='pil'),
+            ]        
+
+    def check_similarity(self, img_urls, model):
+        imgs = []
+        for url in img_urls:
+            if url == "": continue
+            imgs.append(image_util.load_image_url(url, required_size=(model.image_size, model.image_size), image_type=model.image_input_type))
+        
+        features = model.model_cls.extract_feature(imgs)
+        results = []
+        for i, v in enumerate(features):
+            if i == 0: continue 
+            dist = matrix.cosine(features[0], v)
+            print(f'{i} -- distance: {dist}')
+            # results.append((imgs[i], f'similarity: {int(dist*100)}%'))
+            original_img = image_util.load_image_url(img_urls[i], required_size=None, image_type='pil')
+            results.append((original_img, f'similarity: {int(dist*100)}%'))
+
+        return results
+
+    

src/util/image.py

@@ -0,0 +1,13 @@
+from PIL import Image 
+import numpy as np
+import requests
+
+def load_image_url(url, required_size = (224,224), image_type = 'array'):
+    print(f'downloading.. {url}, type: {image_type}')
+    img = Image.open(requests.get(url, stream=True).raw)
+    img = Image.fromarray(np.array(img))
+    if required_size is not None:
+        img = img.resize(required_size)
+    if image_type == 'array':
+        img = (np.expand_dims(np.array(img), 0)/255).astype(np.float32)
+    return img

src/util/matrix.py

@@ -0,0 +1,5 @@
+from numpy.linalg import norm
+import numpy as np
+
+def cosine(x, y):
+    return np.dot(x,y)/(norm(x)*norm(y))