feat(encoding): switching to PyAV for ffmpeg related tasks (#983)

benchmarks/video/run_video_benchmark.py

@@ -416,7 +416,7 @@ if __name__ == "__main__":
         "--vcodec",
         type=str,
         nargs="*",
-        default=["libx264", "libx265", "libsvtav1"],
+        default=["libx264", "hevc", "libsvtav1"],
         help="Video codecs to be tested",
     )
     parser.add_argument(
@@ -446,7 +446,7 @@ if __name__ == "__main__":
     #     nargs="*",
     #     default=[0, 1],
     #     help="Use the fastdecode tuning option. 0 disables it. "
-    #         "For libx264 and libx265, only 1 is possible. "
+    #         "For libx264 and libx265/hevc, only 1 is possible. "
     #         "For libsvtav1, 1, 2 or 3 are possible values with a higher number meaning a faster decoding optimization",
     # )
     parser.add_argument(

lerobot/common/datasets/video_utils.py

@@ -13,16 +13,15 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
+import glob
 import importlib
-import json
 import logging
-import subprocess
 import warnings
-from collections import OrderedDict
 from dataclasses import dataclass, field
 from pathlib import Path
 from typing import Any, ClassVar
 
+import av
 import pyarrow as pa
 import torch
 import torchvision
@@ -252,51 +251,83 @@ def encode_video_frames(
     g: int | None = 2,
     crf: int | None = 30,
     fast_decode: int = 0,
-    log_level: str | None = "error",
+    log_level: int | None = av.logging.ERROR,
     overwrite: bool = False,
 ) -> None:
     """More info on ffmpeg arguments tuning on `benchmark/video/README.md`"""
+    # Check encoder availability
+    if vcodec not in ["h264", "hevc", "libsvtav1"]:
+        raise ValueError(f"Unsupported video codec: {vcodec}. Supported codecs are: h264, hevc, libsvtav1.")
+
     video_path = Path(video_path)
     imgs_dir = Path(imgs_dir)
-    video_path.parent.mkdir(parents=True, exist_ok=True)
 
-    ffmpeg_args = OrderedDict(
+    video_path.parent.mkdir(parents=True, exist_ok=overwrite)
-        [
+
-            ("-f", "image2"),
+    # Encoders/pixel formats incompatibility check
-            ("-r", str(fps)),
+    if (vcodec == "libsvtav1" or vcodec == "hevc") and pix_fmt == "yuv444p":
-            ("-i", str(imgs_dir / "frame_%06d.png")),
+        logging.warning(
-            ("-vcodec", vcodec),
+            f"Incompatible pixel format 'yuv444p' for codec {vcodec}, auto-selecting format 'yuv420p'"
-            ("-pix_fmt", pix_fmt),
+        )
-        ]
+        pix_fmt = "yuv420p"
+
+    # Get input frames
+    template = "frame_" + ("[0-9]" * 6) + ".png"
+    input_list = sorted(
+        glob.glob(str(imgs_dir / template)), key=lambda x: int(x.split("_")[-1].split(".")[0])
     )
 
+    # Define video output frame size (assuming all input frames are the same size)
+    if len(input_list) == 0:
+        raise FileNotFoundError(f"No images found in {imgs_dir}.")
+    dummy_image = Image.open(input_list[0])
+    width, height = dummy_image.size
+
+    # Define video codec options
+    video_options = {}
+
     if g is not None:
-        ffmpeg_args["-g"] = str(g)
+        video_options["g"] = str(g)
 
     if crf is not None:
-        ffmpeg_args["-crf"] = str(crf)
+        video_options["crf"] = str(crf)
 
     if fast_decode:
-        key = "-svtav1-params" if vcodec == "libsvtav1" else "-tune"
+        key = "svtav1-params" if vcodec == "libsvtav1" else "tune"
         value = f"fast-decode={fast_decode}" if vcodec == "libsvtav1" else "fastdecode"
-        ffmpeg_args[key] = value
+        video_options[key] = value
 
+    # Set logging level
     if log_level is not None:
-        ffmpeg_args["-loglevel"] = str(log_level)
+        # "While less efficient, it is generally preferable to modify logging with Python’s logging"
+        logging.getLogger("libav").setLevel(log_level)
 
-    ffmpeg_args = [item for pair in ffmpeg_args.items() for item in pair]
+    # Create and open output file (overwrite by default)
-    if overwrite:
+    with av.open(str(video_path), "w") as output:
-        ffmpeg_args.append("-y")
+        output_stream = output.add_stream(vcodec, fps, options=video_options)
+        output_stream.pix_fmt = pix_fmt
+        output_stream.width = width
+        output_stream.height = height
 
-    ffmpeg_cmd = ["ffmpeg"] + ffmpeg_args + [str(video_path)]
+        # Loop through input frames and encode them
-    # redirect stdin to subprocess.DEVNULL to prevent reading random keyboard inputs from terminal
+        for input_data in input_list:
-    subprocess.run(ffmpeg_cmd, check=True, stdin=subprocess.DEVNULL)
+            input_image = Image.open(input_data).convert("RGB")
+            input_frame = av.VideoFrame.from_image(input_image)
+            packet = output_stream.encode(input_frame)
+            if packet:
+                output.mux(packet)
+
+        # Flush the encoder
+        packet = output_stream.encode()
+        if packet:
+            output.mux(packet)
+
+    # Reset logging level
+    if log_level is not None:
+        av.logging.restore_default_callback()
 
     if not video_path.exists():
-        raise OSError(
+        raise OSError(f"Video encoding did not work. File not found: {video_path}.")
-            f"Video encoding did not work. File not found: {video_path}. "
-            f"Try running the command manually to debug: `{''.join(ffmpeg_cmd)}`"
-        )
 
 
 @dataclass
@@ -332,78 +363,68 @@ with warnings.catch_warnings():
 
 
 def get_audio_info(video_path: Path | str) -> dict:
-    ffprobe_audio_cmd = [
+    # Set logging level
-        "ffprobe",
+    logging.getLogger("libav").setLevel(av.logging.ERROR)
-        "-v",
-        "error",
-        "-select_streams",
-        "a:0",
-        "-show_entries",
-        "stream=channels,codec_name,bit_rate,sample_rate,bit_depth,channel_layout,duration",
-        "-of",
-        "json",
-        str(video_path),
-    ]
-    result = subprocess.run(ffprobe_audio_cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
-    if result.returncode != 0:
-        raise RuntimeError(f"Error running ffprobe: {result.stderr}")
 
-    info = json.loads(result.stdout)
+    # Getting audio stream information
-    audio_stream_info = info["streams"][0] if info.get("streams") else None
+    audio_info = {}
-    if audio_stream_info is None:
+    with av.open(str(video_path), "r") as audio_file:
-        return {"has_audio": False}
+        try:
+            audio_stream = audio_file.streams.audio[0]
+        except IndexError:
+            # Reset logging level
+            av.logging.restore_default_callback()
+            return {"has_audio": False}
 
-    # Return the information, defaulting to None if no audio stream is present
+        audio_info["audio.channels"] = audio_stream.channels
-    return {
+        audio_info["audio.codec"] = audio_stream.codec.canonical_name
-        "has_audio": True,
+        # In an ideal loseless case : bit depth x sample rate x channels = bit rate.
-        "audio.channels": audio_stream_info.get("channels", None),
+        # In an actual compressed case, the bit rate is set according to the compression level : the lower the bit rate, the more compression is applied.
-        "audio.codec": audio_stream_info.get("codec_name", None),
+        audio_info["audio.bit_rate"] = audio_stream.bit_rate
-        "audio.bit_rate": int(audio_stream_info["bit_rate"]) if audio_stream_info.get("bit_rate") else None,
+        audio_info["audio.sample_rate"] = audio_stream.sample_rate  # Number of samples per second
-        "audio.sample_rate": int(audio_stream_info["sample_rate"])
+        # In an ideal loseless case : fixed number of bits per sample.
-        if audio_stream_info.get("sample_rate")
+        # In an actual compressed case : variable number of bits per sample (often reduced to match a given depth rate).
-        else None,
+        audio_info["audio.bit_depth"] = audio_stream.format.bits
-        "audio.bit_depth": audio_stream_info.get("bit_depth", None),
+        audio_info["audio.channel_layout"] = audio_stream.layout.name
-        "audio.channel_layout": audio_stream_info.get("channel_layout", None),
+        audio_info["has_audio"] = True
-    }
+
+    # Reset logging level
+    av.logging.restore_default_callback()
+
+    return audio_info
 
 
 def get_video_info(video_path: Path | str) -> dict:
-    ffprobe_video_cmd = [
+    # Set logging level
-        "ffprobe",
+    logging.getLogger("libav").setLevel(av.logging.ERROR)
-        "-v",
-        "error",
-        "-select_streams",
-        "v:0",
-        "-show_entries",
-        "stream=r_frame_rate,width,height,codec_name,nb_frames,duration,pix_fmt",
-        "-of",
-        "json",
-        str(video_path),
-    ]
-    result = subprocess.run(ffprobe_video_cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
-    if result.returncode != 0:
-        raise RuntimeError(f"Error running ffprobe: {result.stderr}")
 
-    info = json.loads(result.stdout)
+    # Getting video stream information
-    video_stream_info = info["streams"][0]
+    video_info = {}
+    with av.open(str(video_path), "r") as video_file:
+        try:
+            video_stream = video_file.streams.video[0]
+        except IndexError:
+            # Reset logging level
+            av.logging.restore_default_callback()
+            return {}
 
-    # Calculate fps from r_frame_rate
+        video_info["video.height"] = video_stream.height
-    r_frame_rate = video_stream_info["r_frame_rate"]
+        video_info["video.width"] = video_stream.width
-    num, denom = map(int, r_frame_rate.split("/"))
+        video_info["video.codec"] = video_stream.codec.canonical_name
-    fps = num / denom
+        video_info["video.pix_fmt"] = video_stream.pix_fmt
+        video_info["video.is_depth_map"] = False
 
-    pixel_channels = get_video_pixel_channels(video_stream_info["pix_fmt"])
+        # Calculate fps from r_frame_rate
+        video_info["video.fps"] = int(video_stream.base_rate)
 
-    video_info = {
+        pixel_channels = get_video_pixel_channels(video_stream.pix_fmt)
-        "video.fps": fps,
+        video_info["video.channels"] = pixel_channels
-        "video.height": video_stream_info["height"],
+
-        "video.width": video_stream_info["width"],
+    # Reset logging level
-        "video.channels": pixel_channels,
+    av.logging.restore_default_callback()
-        "video.codec": video_stream_info["codec_name"],
+
-        "video.pix_fmt": video_stream_info["pix_fmt"],
+    # Adding audio stream information
-        "video.is_depth_map": False,
+    video_info.update(**get_audio_info(video_path))
-        **get_audio_info(video_path),
-    }
 
     return video_info
 

pyproject.toml

@@ -62,7 +62,7 @@ dependencies = [
     "omegaconf>=2.3.0",
     "opencv-python-headless>=4.9.0",
     "packaging>=24.2",
-    "av>=12.0.5",
+    "av>=14.2.0",
     "pymunk>=6.6.0",
     "pynput>=1.7.7",
     "pyzmq>=26.2.1",