Add captions to detected images gallery

app.py

@@ -1,9 +1,9 @@
 from threading import Thread
 
-from PIL import Image
 import gradio as gr
 import torch
-from transformers import PreTrainedModel, AutoImageProcessor  # for type hint
+from PIL import Image
+from transformers import PreTrainedModel  # for type hint
 from transformers import TextIteratorStreamer, AutoModelForCausalLM, AutoTokenizer  # Moondream
 from transformers import YolosImageProcessor, YolosForObjectDetection  # YOLOS-small-300
 
@@ -64,7 +64,10 @@ def detect_objects(img: Image.Image):
             f"Detected {yolos_model.config.id2label[label.item()]} with confidence "
             f"{round(score.item(), 3)} at location {box}"
         )
-        box_images.append(img.crop((box[0], box[1], box[2], box[3])))
+        box_images.append((
+            img.crop((box[0], box[1], box[2], box[3])),
+            f"{yolos_model.config.id2label[label.item()]} ({round(score.item(), 3)})")
+        )
 
     return box_images
 
@@ -82,7 +85,7 @@ if __name__ == "__main__":
         with gr.Tab("Object Detection"):
             with gr.Row():
                 yolos_input = gr.Image(type="pil")
-                yolos_output = gr.Gallery(label="Detected Objects", show_label=True)
+                yolos_output = gr.Gallery(label="Detected Objects", object_fit="scale-down", columns=3)
             yolos_button = gr.Button("Submit")
 
         with gr.Tab("Inference"):