Update app.py

app.py

@@ -1,25 +1,19 @@
-from huggingface_hub import from_pretrained_fastai
 import gradio as gr
-from fastai.vision.all import *
 from icevision.all import *
-from icevision.models.checkpoint import *
 import PIL
 
-checkpoint_path = "fasterRCNNKangaroo.pth"
+class_map = ClassMap(['kangaroo'])
 model = models.torchvision.faster_rcnn.model(backbone=models.torchvision.faster_rcnn.backbones.resnet50_fpn,
-                                             num_classes=2)
-
-state_dict = torch.load(checkpoint_path, map_location=torch.device('cpu'))
+                                             num_classes=len(class_map))
+state_dict = torch.load('fasterRCNNKangaroo.pth')
 model.load_state_dict(state_dict)
 
-infer_tfms = tfms.A.Adapter([*tfms.A.resize_and_pad(384),tfms.A.Normalize()])
-
-# Definimos una función que se encarga de llevar a cabo las predicciones
+infer_tfms = tfms.A.Adapter([*tfms.A.resize_and_pad(size),tfms.A.Normalize()])
+size = 384
 def predict(img):
-    img = PIL.Image.fromarray(img, "RGB")
-    pred_dict =  models.ross.efficientdet.end2end_detect(img, infer_tfms, model.to("cpu"), class_map=ClassMap(['kangaroo']), detection_threshold=0.5)
-    return pred_dict["img"]
-    
+    img = PILImage.create(img)
+    pred_dict  = models.torchvision.faster_rcnn.end2end_detect(img, infer_tfms, model.to("cpu"), class_map=class_map, detection_threshold=0.5)
+    return pred_dict['img']
+
 # Creamos la interfaz y la lanzamos. 
-gr.Interface(fn=predict, inputs=gr.inputs.Image(shape=(128, 128)), outputs=[gr.outputs.Image(type="pil", label="VFNet Inference")],
-             examples=['00004.jpg','00083.jpg', '00119.jpg']).launch(share=False)
+gr.Interface(fn=predict, inputs=gr.inputs.Image(shape=(128, 128)), outputs=gr.outputs.Image(),examples=['00004.jpg','00083.jpg', '00119.jpg']).launch(share=False)