refactor code

app.py

@@ -101,8 +101,9 @@ def setup_gradio_interface(model, device, colors, color_map):
                 run_button = gr.Button(value="Run")
 
             with gr.Column():
-                overlay_mask = gr.Image(type="numpy", label="Overlay Mask")
-                only_mask = gr.Image(type="numpy", label="Segmentation Mask")
+                with gr.Row():
+                    overlay_mask = gr.Image(type="numpy", label="Overlay Mask")
+                    only_mask = gr.Image(type="numpy", label="Segmentation Mask")
                 h_text = gr.HighlightedText(
                     label="Labels",
                     combine_adjacent=False,

app_demo1.py

@@ -1,138 +0,0 @@
-# from models.builder import build_model
-# from visualization import mask2rgb
-# from segmentation.datasets import PascalVOCDataset
-
-# import os
-# from hydra import compose, initialize
-# from PIL import Image
-# import matplotlib.pyplot as plt
-# from torchvision import transforms as T
-# import torch.nn.functional as F
-# import numpy as np
-from operator import itemgetter
-# import torch
-# import random
-# import warnings
-
-warnings.filterwarnings("ignore")
-initialize(config_path="configs", version_base=None)
-
-# from huggingface_hub import Repository
-
-repo = Repository(
-    local_dir="clip-dinoiser",
-    clone_from="ariG23498/clip-dinoiser",
-    use_auth_token=os.environ.get("token")
-)
-
-check_path = 'clip-dinoiser/checkpoints/last.pt'
-device = "cuda" if torch.cuda.is_available() else "cpu"
-
-check = torch.load(check_path, map_location=device)
-dinoclip_cfg = "clip_dinoiser.yaml"
-cfg = compose(config_name=dinoclip_cfg)
-
-model = build_model(cfg.model, class_names=PascalVOCDataset.CLASSES).to(device)
-model.clip_backbone.decode_head.use_templates=False # switching off the imagenet templates for fast inference
-model.load_state_dict(check['model_state_dict'], strict=False)
-model = model.eval()
-
-# import gradio as gr
-
-colors = [
-    (0, 255, 0),
-    (0, 0, 255),
-    (255, 255, 0),
-    (255, 0, 255),
-    (0, 255, 255),
-    (114, 128, 250),
-    (0, 165, 255),
-    (0, 128, 0),
-    (144, 238, 144),
-    (238, 238, 175),
-    (255, 191, 0),
-    (0, 128, 0),
-    (226, 43, 138),
-    (255, 0, 255),
-    (0, 215, 255),
-    (255, 0, 0),
-]
-
-color_map = {
-    f"{color_id}": f"#{hex(color[0])[2:].zfill(2)}{hex(color[1])[2:].zfill(2)}{hex(color[2])[2:].zfill(2)}" for color_id, color in enumerate(colors)
-}
-
-def run_clip_dinoiser(input_image, text_prompts):
-    image = input_image.convert("RGB")
-    text_prompts = text_prompts.split(",")
-    palette = colors[:len(text_prompts)]
-
-    model.clip_backbone.decode_head.update_vocab(text_prompts)
-    model.to(device)
-    model.apply_found = True
-
-    img_tens = T.PILToTensor()(image).unsqueeze(0).to(device) / 255.
-
-    h, w = img_tens.shape[-2:]
-    output = model(img_tens).cpu()
-    output = F.interpolate(output, scale_factor=model.clip_backbone.backbone.patch_size, mode="bilinear",
-                        align_corners=False)[..., :h, :w]
-    output = output[0].argmax(dim=0)
-    mask = mask2rgb(output, palette)
-
-    classes = np.unique(output).tolist()
-    palette_array = np.array(itemgetter(*classes)(palette)).reshape(1, -1, 3)
-    alpha=0.5
-    blend = (alpha)*np.array(image)/255. + (1-alpha) * mask/255.
-
-    h_text = list()
-    for idx, text in enumerate(text_prompts):
-        h_text.append((text, f"{idx}"))
-    return blend, mask, h_text
-
-
-
-
-if __name__ == "__main__":
-
-    block = gr.Blocks().queue()
-    with block:
-        gr.Markdown("<h1><center>CLIP-DINOiser<h1><center>")
-
-        with gr.Row():
-            with gr.Column():
-                input_image = gr.Image(type="pil", label="Input Image")
-                text_prompts = gr.Textbox(label="Enter comma-separated prompts")
-                run_button = gr.Button(value="Run")
-
-            with gr.Column():
-                with gr.Row():
-                    overlay_mask = gr.Image(
-                        type="numpy",
-                        label="Overlay Mask",
-                    )
-                    only_mask = gr.Image(
-                        type="numpy",
-                        label="Segmentation Mask"
-                    )
-                h_text = gr.HighlightedText(
-                    label="Labels",
-                    combine_adjacent=False,
-                    show_legend=False,
-                    color_map=color_map
-                )
-
-        run_button.click(
-            fn=run_clip_dinoiser,
-            inputs=[input_image, text_prompts,],
-            outputs=[overlay_mask, only_mask, h_text]
-        )
-        gr.Examples(
-            [["vintage_bike.jpeg", "background, vintage bike, leather bag"]],
-            inputs = [input_image, text_prompts,],
-            outputs = [overlay_mask, only_mask, h_text],
-            fn=run_clip_dinoiser,
-            cache_examples=True,
-            label='Try this example input!'
-      )
-    block.launch(share=False, show_api=False, show_error=True)