fix(concatenation compatibility): adding compatibility check when concatenating video files

src/lerobot/datasets/aggregate.py

@@ -332,7 +332,6 @@ def aggregate_videos(src_meta, dst_meta, videos_idx, video_files_size_in_mb, chu
         videos_idx: Dictionary tracking video chunk and file indices.
         video_files_size_in_mb: Maximum size for video files in MB (defaults to DEFAULT_VIDEO_FILE_SIZE_IN_MB)
         chunk_size: Maximum number of files per chunk (defaults to DEFAULT_CHUNK_SIZE)
-
     Returns:
         dict: Updated videos_idx with current chunk and file indices.
     """
@@ -417,6 +416,7 @@ def aggregate_videos(src_meta, dst_meta, videos_idx, video_files_size_in_mb, chu
                 concatenate_video_files(
                     [dst_path, src_path],
                     dst_path,
+                    compatibilty_check=True,
                 )
                 # Update duration of this destination file
                 dst_file_durations[dst_key] = current_dst_duration + src_duration

src/lerobot/datasets/video_utils.py

@@ -555,7 +555,7 @@ def encode_video_frames(
 
 
 def concatenate_video_files(
-    input_video_paths: list[Path | str], output_video_path: Path, overwrite: bool = True
+    input_video_paths: list[Path | str], output_video_path: Path, overwrite: bool = True, compatibilty_check: bool = False
 ):
     """
     Concatenate multiple video files into a single video file using pyav.
@@ -568,6 +568,7 @@ def concatenate_video_files(
         input_video_paths: Ordered list of input video file paths to concatenate.
         output_video_path: Path to the output video file.
         overwrite: Whether to overwrite the output video file if it already exists. Default is True.
+        compatibilty_check: Whether to check if the input videos are compatible. Default is False.
 
     Note:
         - Creates a temporary directory for intermediate files that is cleaned up after use.
@@ -586,6 +587,14 @@ def concatenate_video_files(
     if len(input_video_paths) == 0:
         raise FileNotFoundError("No input video paths provided.")
 
+    # This check may be skipped at recording time as videos are encoded with the same encoder config.
+    if compatibilty_check:
+        reference_video_info = get_video_info(input_video_paths[0])
+        for input_path in input_video_paths[1:]:
+            video_info = get_video_info(input_path)
+            if video_info["video.height"] != reference_video_info["video.height"] or video_info["video.width"] != reference_video_info["video.width"] or video_info["video.fps"] != reference_video_info["video.fps"] or video_info["video.codec"] != reference_video_info["video.codec"] or video_info["video.pix_fmt"] != reference_video_info["video.pix_fmt"]:
+                raise ValueError(f"Input video {input_path} is not compatible with the reference video {input_video_paths[0]}.")
+
     # Create a temporary .ffconcat file to list the input video paths
     with tempfile.NamedTemporaryFile(mode="w", suffix=".ffconcat", delete=False) as tmp_concatenate_file:
         tmp_concatenate_file.write("ffconcat version 1.0\n")