Update app.py

app.py

@@ -74,20 +74,19 @@ def convert_seg_to_bbox(api_key: str, dataset_url: str, split_ratios=(0.8, 0.1,
         )
         annos.setdefault(img_id, []).append(line)
 
-    # --- locate the single folder of raw images
-    img_src = None
+    # --- map each file_name to its actual path on disk
+    name_to_id = {img['file_name']: img['id'] for img in coco['images']}
+    file_paths = {}
     for dp, _, files in os.walk(root):
-        if any(f.lower().endswith(('.jpg','.png','.jpeg')) for f in files):
-            img_src = dp
-            break
-    if not img_src:
-        raise FileNotFoundError(f"No images under {root}")
+        for f in files:
+            if f in name_to_id:
+                file_paths[f] = os.path.join(dp, f)
 
-    # --- copy images + write flat label files
-    name_to_id = {img['file_name']: img['id'] for img in coco['images']}
+    # --- copy images + write flat_labels
     for fname, img_id in name_to_id.items():
-        src_path = os.path.join(img_src, fname)
-        if not os.path.isfile(src_path):
+        src_path = file_paths.get(fname)
+        if not src_path:
+            # skip if we couldn't find this image under root
             continue
         shutil.copy(src_path, os.path.join(flat_img, fname))
         with open(os.path.join(flat_lbl, fname.rsplit('.',1)[0] + ".txt"), 'w') as lf:
@@ -102,7 +101,7 @@ def convert_seg_to_bbox(api_key: str, dataset_url: str, split_ratios=(0.8, 0.1,
     n = len(all_files)
     n_train = max(1, int(n * split_ratios[0]))
     n_valid = max(1, int(n * split_ratios[1]))
-    # ensure at least 1 for each split
+    # ensure at least 1 left for test
     n_valid = min(n_valid, n - n_train - 1)
 
     splits = {
@@ -118,12 +117,10 @@ def convert_seg_to_bbox(api_key: str, dataset_url: str, split_ratios=(0.8, 0.1,
         os.makedirs(out_img_dir, exist_ok=True)
         os.makedirs(out_lbl_dir, exist_ok=True)
         for fn in files:
-            # move image
             shutil.move(
                 os.path.join(flat_img, fn),
                 os.path.join(out_img_dir, fn)
             )
-            # move corresponding .txt label
             lbl_fn = fn.rsplit('.',1)[0] + ".txt"
             shutil.move(
                 os.path.join(flat_lbl, lbl_fn),
@@ -138,7 +135,9 @@ def convert_seg_to_bbox(api_key: str, dataset_url: str, split_ratios=(0.8, 0.1,
     before, after = [], []
     sample = random.sample(list(name_to_id.keys()), min(5, len(name_to_id)))
     for fname in sample:
-        src = os.path.join(img_src, fname)
+        src = file_paths.get(fname)
+        if not src:
+            continue
         img = cv2.cvtColor(cv2.imread(src), cv2.COLOR_BGR2RGB)
 
         seg_vis = img.copy()
@@ -174,7 +173,7 @@ def upload_and_train_detection(
     rf = Roboflow(api_key=api_key)
     ws = rf.workspace()
 
-    # get-or-create detection project
+    # get‐or‐create your detection project
     try:
         proj = ws.project(project_slug)
     except:
@@ -185,7 +184,7 @@ def upload_and_train_detection(
             project_license=project_license
         )
 
-    # upload the folder that now has train/valid/test
+    # upload the folder with proper train/valid/test
     ws.upload_dataset(
         dataset_path,
         project_slug,