Update app.py
Browse files
app.py
CHANGED
@@ -1,66 +1,24 @@
|
|
1 |
import gradio as gr
|
2 |
-
import
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
-
#
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
#
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
results = []
|
26 |
-
|
27 |
-
while cap.isOpened():
|
28 |
-
ret, frame = cap.read()
|
29 |
-
if not ret:
|
30 |
-
break
|
31 |
-
|
32 |
-
# Object detection
|
33 |
-
results_detection = model(frame)
|
34 |
-
|
35 |
-
# Logic for determining runner status using detected objects
|
36 |
-
objects = results_detection.pred[0][:, -1].numpy()
|
37 |
-
if 0 in objects: # 0 corresponds to person class
|
38 |
-
# Perform OCR on the detected person
|
39 |
-
person_bbox = results_detection.pred[0][np.where(objects == 0)][0][:4]
|
40 |
-
person_bbox = person_bbox.astype(int)
|
41 |
-
person_img = frame[person_bbox[1]:person_bbox[3], person_bbox[0]:person_bbox[2]]
|
42 |
-
text = perform_ocr(person_img)
|
43 |
-
|
44 |
-
# Classification using text classification model
|
45 |
-
inputs_classification = classification_tokenizer(text, return_tensors="pt", padding=True, truncation=True)
|
46 |
-
outputs_classification = classification_model(**inputs_classification)
|
47 |
-
predicted_class = torch.argmax(outputs_classification.logits).item()
|
48 |
-
if predicted_class == 1:
|
49 |
-
runner_status = "Out"
|
50 |
-
else:
|
51 |
-
runner_status = "Safe"
|
52 |
-
|
53 |
-
result = {
|
54 |
-
"frame_number": cap.get(cv2.CAP_PROP_POS_FRAMES),
|
55 |
-
"runner_status": runner_status
|
56 |
-
}
|
57 |
-
results.append(result)
|
58 |
-
|
59 |
-
cap.release()
|
60 |
-
|
61 |
-
return results
|
62 |
-
|
63 |
-
inputs = gr.inputs.Video(type="file", label="Upload a baseball video")
|
64 |
-
outputs = gr.outputs.Label(type="auto", label="Runner Status")
|
65 |
-
interface = gr.Interface(fn=predict_runner_status, inputs=inputs, outputs=outputs, title="Baseball Runner Status Predictor")
|
66 |
-
interface.launch(share=True)
|
|
|
1 |
import gradio as gr
|
2 |
+
from transformers import pipeline
|
3 |
+
|
4 |
+
# Replace with a suitable image classification model ID
|
5 |
+
model_id = "sayakpaul/resnet-50-finetuned-imagenet"
|
6 |
+
|
7 |
+
def analyze_image(image):
|
8 |
+
classifier = pipeline("image-classification", model=model_id)
|
9 |
+
predictions = classifier(images=image) # Assuming the model outputs probabilities
|
10 |
+
# Extract the most likely class and its probability
|
11 |
+
top_class = predictions[0]["label"]
|
12 |
+
top_prob = predictions[0]["score"]
|
13 |
+
return f"Top Class: {top_class} (Probability: {top_prob:.2f})"
|
14 |
+
|
15 |
+
# Gradio interface
|
16 |
+
interface = gr.Interface(
|
17 |
+
fn=analyze_image,
|
18 |
+
inputs="image",
|
19 |
+
outputs="text",
|
20 |
+
title="Image Analyzer (Generic)",
|
21 |
+
description="Upload an image and get the most likely classification based on the chosen model.",
|
22 |
+
)
|
23 |
+
|
24 |
+
interface.launch()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|