Deploy Streamlit app to Hugging Face Space

README.md

@@ -1,20 +1,36 @@
+# 🎨 Image Colorization & Post-Processing Tool
+
+This project provides a **Streamlit-based web app** for automatic image **colorization** and **basic post-processing** (sharpening, blurring, undo, saving). It's built using a pretrained deep learning model from [Zhang et al. (2017)](https://richzhang.github.io/colorization/).
+
 ---
-title: Colorization
-emoji: 🚀
-colorFrom: red
-colorTo: red
-sdk: docker
-app_port: 8501
-tags:
-- streamlit
-pinned: false
-short_description: A Streamlit app that colorizes black-and-white images
-license: apache-2.0
+
+## 🚀 Features
+
+- ✅ Upload grayscale image
+- 🎨 Auto-colorize using `siggraph17` model
+- 🧪 Sharpen or blur the result
+- ↩️ Undo changes (1-step)
+- 💾 Save processed image
+
+---
+
+## 🛠️ Technologies Used
+
+- Python
+- Streamlit
+- PyTorch
+- OpenCV
+- PIL (Pillow)
+- `colorizers` (Zhang's pretrained models)
+
 ---
 
-# Welcome to Streamlit!
+## 📦 Installation
 
-Edit `/src/streamlit_app.py` to customize this app to your heart's desire. :heart:
+```bash
+# Clone the repository
+git clone https://github.com/haiderakt/image-colorization.git
+cd image-colorization
 
-If you have any questions, checkout our [documentation](https://docs.streamlit.io) and [community
-forums](https://discuss.streamlit.io).
+# Install dependencies
+pip install -r requirements.txt

app.py

@@ -0,0 +1,92 @@
+import cv2
+import numpy as np
+import torch
+import streamlit as st
+from PIL import Image
+import colorizers
+
+# Load pretrained colorization model
+model = colorizers.siggraph17(pretrained=True).eval()
+
+# Session state init
+if 'processed_image' not in st.session_state:
+    st.session_state.processed_image = None
+if 'original_image' not in st.session_state:
+    st.session_state.original_image = None
+if 'history' not in st.session_state:
+    st.session_state.history = []
+
+# Convert OpenCV image to PIL
+def display_image_cv2(image):
+    rgb_img = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
+    return Image.fromarray(rgb_img)
+
+# Colorization logic
+def colouring_image(file, model):
+    img = cv2.imdecode(np.frombuffer(file.read(), np.uint8), cv2.IMREAD_GRAYSCALE)
+    original = cv2.cvtColor(cv2.resize(img, (256, 256)), cv2.COLOR_GRAY2BGR)
+
+    img = cv2.resize(img, (256, 256)) / 255.0 * 100
+    img_tensor = torch.from_numpy(img).unsqueeze(0).unsqueeze(0).float()
+
+    with torch.no_grad():
+        ab = model(img_tensor).cpu().numpy()[0].transpose((1, 2, 0))
+
+    lab = np.concatenate((img[:, :, np.newaxis], ab), axis=2)
+    bgr = cv2.cvtColor(lab.astype(np.float32), cv2.COLOR_Lab2BGR)
+    bgr = np.clip(bgr * 255, 0, 255).astype(np.uint8)
+
+    return bgr, original
+
+# UI Setup
+st.set_page_config(page_title="Image Colorizer", layout="wide")
+st.title("🎨 Image Colorization and Post-Processing Tool")
+
+uploaded_file = st.file_uploader("Upload a grayscale image", type=["jpg", "jpeg", "png", "bmp"])
+
+if uploaded_file:
+    colorized, original = colouring_image(uploaded_file, model)
+    st.session_state.processed_image = colorized.copy()
+    st.session_state.original_image = original
+    st.session_state.history = [colorized.copy()]
+
+    st.subheader("Preview:")
+    col1, col2 = st.columns(2)
+    with col1:
+        st.image(display_image_cv2(original), caption="Original Image", use_container_width=True)
+    with col2:
+        st.image(display_image_cv2(colorized), caption="Colorized Image", use_container_width=True)
+
+    st.markdown("---")
+
+    # Button row
+    colA, colB, colC, colD = st.columns(4)
+
+    with colA:
+        if st.button("🔪 Sharpen"):
+            kernel = np.array([[0, -1, 0], [-1, 5, -1], [0, -1, 0]])
+            sharpened = cv2.filter2D(st.session_state.processed_image, -1, kernel)
+            st.session_state.history.append(st.session_state.processed_image.copy())
+            st.session_state.processed_image = sharpened
+            st.image(display_image_cv2(sharpened), caption="Sharpened Image", use_container_width=True)
+
+    with colB:
+        if st.button("💧 Blur"):
+            blurred = cv2.GaussianBlur(st.session_state.processed_image, (15, 15), 0)
+            st.session_state.history.append(st.session_state.processed_image.copy())
+            st.session_state.processed_image = blurred
+            st.image(display_image_cv2(blurred), caption="Blurred Image", use_container_width=True)
+
+    with colC:
+        if st.button("↩️ Undo"):
+            if len(st.session_state.history) > 1:
+                st.session_state.history.pop()
+                st.session_state.processed_image = st.session_state.history[-1]
+                st.image(display_image_cv2(st.session_state.processed_image), caption="Undo Applied", use_container_width=True)
+            else:
+                st.warning("Nothing to undo.")
+
+    with colD:
+        if st.session_state.processed_image is not None:
+            buf = cv2.imencode(".jpg", st.session_state.processed_image)[1].tobytes()
+            st.download_button(label="💾 Save Image", data=buf, file_name="colorized.jpg", mime="image/jpeg")

colorizers/__init__.py

@@ -0,0 +1,6 @@
+
+from .base_color import *
+from .eccv16 import *
+from .siggraph17 import *
+from .util import *
+

colorizers/base_color.py

@@ -0,0 +1,24 @@
+
+import torch
+from torch import nn
+
+class BaseColor(nn.Module):
+	def __init__(self):
+		super(BaseColor, self).__init__()
+
+		self.l_cent = 50.
+		self.l_norm = 100.
+		self.ab_norm = 110.
+
+	def normalize_l(self, in_l):
+		return (in_l-self.l_cent)/self.l_norm
+
+	def unnormalize_l(self, in_l):
+		return in_l*self.l_norm + self.l_cent
+
+	def normalize_ab(self, in_ab):
+		return in_ab/self.ab_norm
+
+	def unnormalize_ab(self, in_ab):
+		return in_ab*self.ab_norm
+

colorizers/eccv16.py

@@ -0,0 +1,105 @@
+
+import torch
+import torch.nn as nn
+import numpy as np
+from IPython import embed
+
+from .base_color import *
+
+class ECCVGenerator(BaseColor):
+    def __init__(self, norm_layer=nn.BatchNorm2d):
+        super(ECCVGenerator, self).__init__()
+
+        model1=[nn.Conv2d(1, 64, kernel_size=3, stride=1, padding=1, bias=True),]
+        model1+=[nn.ReLU(True),]
+        model1+=[nn.Conv2d(64, 64, kernel_size=3, stride=2, padding=1, bias=True),]
+        model1+=[nn.ReLU(True),]
+        model1+=[norm_layer(64),]
+
+        model2=[nn.Conv2d(64, 128, kernel_size=3, stride=1, padding=1, bias=True),]
+        model2+=[nn.ReLU(True),]
+        model2+=[nn.Conv2d(128, 128, kernel_size=3, stride=2, padding=1, bias=True),]
+        model2+=[nn.ReLU(True),]
+        model2+=[norm_layer(128),]
+
+        model3=[nn.Conv2d(128, 256, kernel_size=3, stride=1, padding=1, bias=True),]
+        model3+=[nn.ReLU(True),]
+        model3+=[nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1, bias=True),]
+        model3+=[nn.ReLU(True),]
+        model3+=[nn.Conv2d(256, 256, kernel_size=3, stride=2, padding=1, bias=True),]
+        model3+=[nn.ReLU(True),]
+        model3+=[norm_layer(256),]
+
+        model4=[nn.Conv2d(256, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model4+=[nn.ReLU(True),]
+        model4+=[nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model4+=[nn.ReLU(True),]
+        model4+=[nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model4+=[nn.ReLU(True),]
+        model4+=[norm_layer(512),]
+
+        model5=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model5+=[nn.ReLU(True),]
+        model5+=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model5+=[nn.ReLU(True),]
+        model5+=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model5+=[nn.ReLU(True),]
+        model5+=[norm_layer(512),]
+
+        model6=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model6+=[nn.ReLU(True),]
+        model6+=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model6+=[nn.ReLU(True),]
+        model6+=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model6+=[nn.ReLU(True),]
+        model6+=[norm_layer(512),]
+
+        model7=[nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model7+=[nn.ReLU(True),]
+        model7+=[nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model7+=[nn.ReLU(True),]
+        model7+=[nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model7+=[nn.ReLU(True),]
+        model7+=[norm_layer(512),]
+
+        model8=[nn.ConvTranspose2d(512, 256, kernel_size=4, stride=2, padding=1, bias=True),]
+        model8+=[nn.ReLU(True),]
+        model8+=[nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1, bias=True),]
+        model8+=[nn.ReLU(True),]
+        model8+=[nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1, bias=True),]
+        model8+=[nn.ReLU(True),]
+
+        model8+=[nn.Conv2d(256, 313, kernel_size=1, stride=1, padding=0, bias=True),]
+
+        self.model1 = nn.Sequential(*model1)
+        self.model2 = nn.Sequential(*model2)
+        self.model3 = nn.Sequential(*model3)
+        self.model4 = nn.Sequential(*model4)
+        self.model5 = nn.Sequential(*model5)
+        self.model6 = nn.Sequential(*model6)
+        self.model7 = nn.Sequential(*model7)
+        self.model8 = nn.Sequential(*model8)
+
+        self.softmax = nn.Softmax(dim=1)
+        self.model_out = nn.Conv2d(313, 2, kernel_size=1, padding=0, dilation=1, stride=1, bias=False)
+        self.upsample4 = nn.Upsample(scale_factor=4, mode='bilinear')
+
+    def forward(self, input_l):
+        conv1_2 = self.model1(self.normalize_l(input_l))
+        conv2_2 = self.model2(conv1_2)
+        conv3_3 = self.model3(conv2_2)
+        conv4_3 = self.model4(conv3_3)
+        conv5_3 = self.model5(conv4_3)
+        conv6_3 = self.model6(conv5_3)
+        conv7_3 = self.model7(conv6_3)
+        conv8_3 = self.model8(conv7_3)
+        out_reg = self.model_out(self.softmax(conv8_3))
+
+        return self.unnormalize_ab(self.upsample4(out_reg))
+
+def eccv16(pretrained=True):
+	model = ECCVGenerator()
+	if(pretrained):
+		import torch.utils.model_zoo as model_zoo
+		model.load_state_dict(model_zoo.load_url('https://colorizers.s3.us-east-2.amazonaws.com/colorization_release_v2-9b330a0b.pth',map_location='cpu',check_hash=True))
+	return model

colorizers/siggraph17.py

@@ -0,0 +1,168 @@
+import torch
+import torch.nn as nn
+
+from .base_color import *
+
+class SIGGRAPHGenerator(BaseColor):
+    def __init__(self, norm_layer=nn.BatchNorm2d, classes=529):
+        super(SIGGRAPHGenerator, self).__init__()
+
+        # Conv1
+        model1=[nn.Conv2d(4, 64, kernel_size=3, stride=1, padding=1, bias=True),]
+        model1+=[nn.ReLU(True),]
+        model1+=[nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=1, bias=True),]
+        model1+=[nn.ReLU(True),]
+        model1+=[norm_layer(64),]
+        # add a subsampling operation
+
+        # Conv2
+        model2=[nn.Conv2d(64, 128, kernel_size=3, stride=1, padding=1, bias=True),]
+        model2+=[nn.ReLU(True),]
+        model2+=[nn.Conv2d(128, 128, kernel_size=3, stride=1, padding=1, bias=True),]
+        model2+=[nn.ReLU(True),]
+        model2+=[norm_layer(128),]
+        # add a subsampling layer operation
+
+        # Conv3
+        model3=[nn.Conv2d(128, 256, kernel_size=3, stride=1, padding=1, bias=True),]
+        model3+=[nn.ReLU(True),]
+        model3+=[nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1, bias=True),]
+        model3+=[nn.ReLU(True),]
+        model3+=[nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1, bias=True),]
+        model3+=[nn.ReLU(True),]
+        model3+=[norm_layer(256),]
+        # add a subsampling layer operation
+
+        # Conv4
+        model4=[nn.Conv2d(256, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model4+=[nn.ReLU(True),]
+        model4+=[nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model4+=[nn.ReLU(True),]
+        model4+=[nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model4+=[nn.ReLU(True),]
+        model4+=[norm_layer(512),]
+
+        # Conv5
+        model5=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model5+=[nn.ReLU(True),]
+        model5+=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model5+=[nn.ReLU(True),]
+        model5+=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model5+=[nn.ReLU(True),]
+        model5+=[norm_layer(512),]
+
+        # Conv6
+        model6=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model6+=[nn.ReLU(True),]
+        model6+=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model6+=[nn.ReLU(True),]
+        model6+=[nn.Conv2d(512, 512, kernel_size=3, dilation=2, stride=1, padding=2, bias=True),]
+        model6+=[nn.ReLU(True),]
+        model6+=[norm_layer(512),]
+
+        # Conv7
+        model7=[nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model7+=[nn.ReLU(True),]
+        model7+=[nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model7+=[nn.ReLU(True),]
+        model7+=[nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1, bias=True),]
+        model7+=[nn.ReLU(True),]
+        model7+=[norm_layer(512),]
+
+        # Conv7
+        model8up=[nn.ConvTranspose2d(512, 256, kernel_size=4, stride=2, padding=1, bias=True)]
+        model3short8=[nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1, bias=True),]
+
+        model8=[nn.ReLU(True),]
+        model8+=[nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1, bias=True),]
+        model8+=[nn.ReLU(True),]
+        model8+=[nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1, bias=True),]
+        model8+=[nn.ReLU(True),]
+        model8+=[norm_layer(256),]
+
+        # Conv9
+        model9up=[nn.ConvTranspose2d(256, 128, kernel_size=4, stride=2, padding=1, bias=True),]
+        model2short9=[nn.Conv2d(128, 128, kernel_size=3, stride=1, padding=1, bias=True),]
+        # add the two feature maps above        
+
+        model9=[nn.ReLU(True),]
+        model9+=[nn.Conv2d(128, 128, kernel_size=3, stride=1, padding=1, bias=True),]
+        model9+=[nn.ReLU(True),]
+        model9+=[norm_layer(128),]
+
+        # Conv10
+        model10up=[nn.ConvTranspose2d(128, 128, kernel_size=4, stride=2, padding=1, bias=True),]
+        model1short10=[nn.Conv2d(64, 128, kernel_size=3, stride=1, padding=1, bias=True),]
+        # add the two feature maps above
+
+        model10=[nn.ReLU(True),]
+        model10+=[nn.Conv2d(128, 128, kernel_size=3, dilation=1, stride=1, padding=1, bias=True),]
+        model10+=[nn.LeakyReLU(negative_slope=.2),]
+
+        # classification output
+        model_class=[nn.Conv2d(256, classes, kernel_size=1, padding=0, dilation=1, stride=1, bias=True),]
+
+        # regression output
+        model_out=[nn.Conv2d(128, 2, kernel_size=1, padding=0, dilation=1, stride=1, bias=True),]
+        model_out+=[nn.Tanh()]
+
+        self.model1 = nn.Sequential(*model1)
+        self.model2 = nn.Sequential(*model2)
+        self.model3 = nn.Sequential(*model3)
+        self.model4 = nn.Sequential(*model4)
+        self.model5 = nn.Sequential(*model5)
+        self.model6 = nn.Sequential(*model6)
+        self.model7 = nn.Sequential(*model7)
+        self.model8up = nn.Sequential(*model8up)
+        self.model8 = nn.Sequential(*model8)
+        self.model9up = nn.Sequential(*model9up)
+        self.model9 = nn.Sequential(*model9)
+        self.model10up = nn.Sequential(*model10up)
+        self.model10 = nn.Sequential(*model10)
+        self.model3short8 = nn.Sequential(*model3short8)
+        self.model2short9 = nn.Sequential(*model2short9)
+        self.model1short10 = nn.Sequential(*model1short10)
+
+        self.model_class = nn.Sequential(*model_class)
+        self.model_out = nn.Sequential(*model_out)
+
+        self.upsample4 = nn.Sequential(*[nn.Upsample(scale_factor=4, mode='bilinear'),])
+        self.softmax = nn.Sequential(*[nn.Softmax(dim=1),])
+
+    def forward(self, input_A, input_B=None, mask_B=None):
+        if(input_B is None):
+            input_B = torch.cat((input_A*0, input_A*0), dim=1)
+        if(mask_B is None):
+            mask_B = input_A*0
+
+        conv1_2 = self.model1(torch.cat((self.normalize_l(input_A),self.normalize_ab(input_B),mask_B),dim=1))
+        conv2_2 = self.model2(conv1_2[:,:,::2,::2])
+        conv3_3 = self.model3(conv2_2[:,:,::2,::2])
+        conv4_3 = self.model4(conv3_3[:,:,::2,::2])
+        conv5_3 = self.model5(conv4_3)
+        conv6_3 = self.model6(conv5_3)
+        conv7_3 = self.model7(conv6_3)
+
+        conv8_up = self.model8up(conv7_3) + self.model3short8(conv3_3)
+        conv8_3 = self.model8(conv8_up)
+        conv9_up = self.model9up(conv8_3) + self.model2short9(conv2_2)
+        conv9_3 = self.model9(conv9_up)
+        conv10_up = self.model10up(conv9_3) + self.model1short10(conv1_2)
+        conv10_2 = self.model10(conv10_up)
+        out_reg = self.model_out(conv10_2)
+
+        conv9_up = self.model9up(conv8_3) + self.model2short9(conv2_2)
+        conv9_3 = self.model9(conv9_up)
+        conv10_up = self.model10up(conv9_3) + self.model1short10(conv1_2)
+        conv10_2 = self.model10(conv10_up)
+        out_reg = self.model_out(conv10_2)
+
+        return self.unnormalize_ab(out_reg)
+
+def siggraph17(pretrained=True):
+    model = SIGGRAPHGenerator()
+    if(pretrained):
+        import torch.utils.model_zoo as model_zoo
+        model.load_state_dict(model_zoo.load_url('https://colorizers.s3.us-east-2.amazonaws.com/siggraph17-df00044c.pth',map_location='cpu',check_hash=True))
+    return model
+

colorizers/util.py

@@ -0,0 +1,47 @@
+
+from PIL import Image
+import numpy as np
+from skimage import color
+import torch
+import torch.nn.functional as F
+from IPython import embed
+
+def load_img(img_path):
+	out_np = np.asarray(Image.open(img_path))
+	if(out_np.ndim==2):
+		out_np = np.tile(out_np[:,:,None],3)
+	return out_np
+
+def resize_img(img, HW=(256,256), resample=3):
+	return np.asarray(Image.fromarray(img).resize((HW[1],HW[0]), resample=resample))
+
+def preprocess_img(img_rgb_orig, HW=(256,256), resample=3):
+	# return original size L and resized L as torch Tensors
+	img_rgb_rs = resize_img(img_rgb_orig, HW=HW, resample=resample)
+	
+	img_lab_orig = color.rgb2lab(img_rgb_orig)
+	img_lab_rs = color.rgb2lab(img_rgb_rs)
+
+	img_l_orig = img_lab_orig[:,:,0]
+	img_l_rs = img_lab_rs[:,:,0]
+
+	tens_orig_l = torch.Tensor(img_l_orig)[None,None,:,:]
+	tens_rs_l = torch.Tensor(img_l_rs)[None,None,:,:]
+
+	return (tens_orig_l, tens_rs_l)
+
+def postprocess_tens(tens_orig_l, out_ab, mode='bilinear'):
+	# tens_orig_l 	1 x 1 x H_orig x W_orig
+	# out_ab 		1 x 2 x H x W
+
+	HW_orig = tens_orig_l.shape[2:]
+	HW = out_ab.shape[2:]
+
+	# call resize function if needed
+	if(HW_orig[0]!=HW[0] or HW_orig[1]!=HW[1]):
+		out_ab_orig = F.interpolate(out_ab, size=HW_orig, mode='bilinear')
+	else:
+		out_ab_orig = out_ab
+
+	out_lab_orig = torch.cat((tens_orig_l, out_ab_orig), dim=1)
+	return color.lab2rgb(out_lab_orig.data.cpu().numpy()[0,...].transpose((1,2,0)))

requirements.txt

@@ -1,3 +1,7 @@
-altair
-pandas
+torch
+scikit-image
+numpy
+matplotlib
+argparse
+Pillow
 streamlit