2424

app.py

@@ -3,7 +3,7 @@ from visualization import visualization
 # pipeline = pipeline(task="image-classification", model="julien-c/hotdog-not-hotdog")
 # pipeline = pipeline(task="image-classification", model="jhp/hoi")
 
-def predict(image,threshold,topk):
+def predict(image,threshold,topk,device=''):
     vis_img = visualization(image,threshold,topk)
     return vis_img
 

gradio.ipynb

@@ -6,12 +6,20 @@
    "id": "531487e5-d72d-41be-b4ae-ccd9f8dc844e",
    "metadata": {},
    "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/tmp/ipykernel_4031598/48305459.py:16: GradioDeprecationWarning: Usage of gradio.outputs is deprecated, and will not be supported in the future, please import your components from gradio.components\n",
+      "  outputs= gr.outputs.Image(type=\"pil\", label=\"hoi detection results\"),\n"
+     ]
+    },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
       "Running on local URL:  http://127.0.0.1:7860\n",
-      "Running on public URL: https://fc8effa414b728bb78.gradio.live\n",
+      "Running on public URL: https://fd9d0145926e3bdb6d.gradio.live\n",
       "\n",
       "This share link expires in 72 hours. For free permanent hosting and GPU upgrades, run `gradio deploy` from Terminal to deploy to Spaces (https://huggingface.co/spaces)\n"
      ]
@@ -19,7 +27,7 @@
     {
      "data": {
       "text/html": [
-       "<div><iframe src=\"https://fc8effa414b728bb78.gradio.live\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"
+       "<div><iframe src=\"https://fd9d0145926e3bdb6d.gradio.live\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"
       ],
       "text/plain": [
        "<IPython.core.display.HTML object>"
@@ -33,7 +41,7 @@
      "output_type": "stream",
      "text": [
       "loading annotations into memory...\n",
-      "Done (t=1.67s)\n",
+      "Done (t=1.56s)\n",
       "creating index...\n",
       "index created!\n",
       "\n",
@@ -77,8 +85,8 @@
     "    predict,\n",
     "    inputs=[gr.Image(type='pil',label=\"input image\"),\n",
     "           gr.Slider(0, 1, value=0.4, label=\"Threshold\", info=\"Set detection score threshold between 0~1\"),\n",
-    "           gr.Number(value=5,label='Topk',info='Topk prediction')],\n",
-    "    outputs= gr.Image(type=\"pil\", label=\"hoi detection results\"),\n",
+    "           gr.Number(value=5,info='Topk prediction')],\n",
+    "    outputs= gr.outputs.Image(type=\"pil\", label=\"hoi detection results\"),\n",
     "    title=\"HOI detection\",\n",
     ").launch(share=True,debug=True)"
    ]

visualization.py

@@ -144,9 +144,11 @@ def vis(args,input_img=None,id=294,return_img=False):
 
     vis_img=draw_img_vcoco(image,output_i,top_k=args.topk,threshold=args.threshold,color=builtin_meta.COCO_CATEGORIES)        
     plt.imshow(cv2.cvtColor(vis_img,cv2.COLOR_BGR2RGB))
-    # import pdb;pdb.set_trace()
+    
     if return_img:
-        return Image.fromarray(vis_img)
+        vis_img = Image.fromarray(vis_img[:,:,::-1])
+        # import pdb;pdb.set_trace()
+        return vis_img
     else:
         cv2.imwrite('./vis_res/vis1.jpg',vis_img)
     
@@ -235,7 +237,7 @@ def visualization(input_img,threshold,topk):
     # args.topk = topk
     if args.output_dir:
         Path(args.output_dir).mkdir(parents=True, exist_ok=True)
-    vis(args,input_img=input_img,return_img=True)
+    return vis(args,input_img=input_img,return_img=True)
 
 if __name__ == '__main__':
     parser = argparse.ArgumentParser('DETR training and evaluation script', parents=[get_args_parser()])