syncnet-python 0.1.0__tar.gz → 0.1.1__tar.gz

{syncnet_python-0.1.0 → syncnet_python-0.1.1}/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.4
 Name: syncnet-python
-Version: 0.1.0
-Summary: SyncNet: Audio-visual synchronization detection using deep learning
+Version: 0.1.1
+Summary: SyncNet: Audio-visual synchronization detection using deep learning. Updated version of https://github.com/joonson/syncnet_python for modern Python versions.
 Author: SyncNet Python Contributors
 Maintainer: SyncNet Python Contributors
 License: MIT
@@ -49,6 +49,8 @@ Dynamic: license-file
 
 Audio-visual synchronization detection using deep learning.
 
+This is an updated version of the original [SyncNet implementation](https://github.com/joonson/syncnet_python) by Joon Son Chung, compatible with modern Python versions (3.9+).
+
 ## Overview
 
 SyncNet Python is a PyTorch implementation of the SyncNet model, which detects audio-visual synchronization in videos. It can identify lip-sync errors by analyzing the correspondence between mouth movements and spoken audio.
@@ -126,9 +128,13 @@ syncnet-python video.mp4 --device cpu
 - CUDA (optional but recommended)
 - FFmpeg
 
+## Credits
+
+This package is based on the original [SyncNet implementation](https://github.com/joonson/syncnet_python) by Joon Son Chung.
+
 ## Citation
 
-If you use this code in your research, please cite:
+If you use this code in your research, please cite the original paper:
 
 ```bibtex
 @inproceedings{chung2016out,

{syncnet_python-0.1.0 → syncnet_python-0.1.1}/README.md

@@ -2,6 +2,8 @@
 
 Audio-visual synchronization detection using deep learning.
 
+This is an updated version of the original [SyncNet implementation](https://github.com/joonson/syncnet_python) by Joon Son Chung, compatible with modern Python versions (3.9+).
+
 ## Overview
 
 SyncNet Python is a PyTorch implementation of the SyncNet model, which detects audio-visual synchronization in videos. It can identify lip-sync errors by analyzing the correspondence between mouth movements and spoken audio.
@@ -79,9 +81,13 @@ syncnet-python video.mp4 --device cpu
 - CUDA (optional but recommended)
 - FFmpeg
 
+## Credits
+
+This package is based on the original [SyncNet implementation](https://github.com/joonson/syncnet_python) by Joon Son Chung.
+
 ## Citation
 
-If you use this code in your research, please cite:
+If you use this code in your research, please cite the original paper:
 
 ```bibtex
 @inproceedings{chung2016out,

{syncnet_python-0.1.0 → syncnet_python-0.1.1}/pyproject.toml

@@ -4,8 +4,8 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "syncnet-python"
-version = "0.1.0"
-description = "SyncNet: Audio-visual synchronization detection using deep learning"
+version = "0.1.1"
+description = "SyncNet: Audio-visual synchronization detection using deep learning. Updated version of https://github.com/joonson/syncnet_python for modern Python versions."
 readme = "README.md"
 requires-python = ">=3.9"
 license = {text = "MIT"}

syncnet_python-0.1.1/syncnet/core/types.py

@@ -0,0 +1,67 @@
+"""Type definitions for SyncNet components."""
+
+from pathlib import Path
+from typing import TypedDict
+try:
+    from typing import TypeAlias, NotRequired
+except ImportError:
+    # Python 3.9 compatibility
+    from typing import Dict, Tuple, List
+    TypeAlias = type
+    NotRequired = lambda x: x
+import numpy as np
+import torch
+
+# Basic type aliases
+if 'TypeAlias' in globals() and TypeAlias != type:
+    BBox: TypeAlias = tuple[float, float, float, float]
+    Frame: TypeAlias = np.ndarray
+    AudioData: TypeAlias = np.ndarray
+    MFCCFeatures: TypeAlias = np.ndarray
+else:
+    # Python 3.9 compatibility
+    BBox = Tuple[float, float, float, float]
+    Frame = np.ndarray
+    AudioData = np.ndarray
+    MFCCFeatures = np.ndarray
+
+# Detection types
+class Detection(TypedDict):
+    """Face detection result."""
+    frame_idx: int
+    bbox: BBox
+    confidence: float
+    landmarks: NotRequired[np.ndarray]
+
+# Track types
+class Track(TypedDict):
+    """Face track information."""
+    frame: np.ndarray
+    bbox: np.ndarray
+    start_frame: int
+    end_frame: int
+
+# Result types
+class SyncResult(TypedDict):
+    """Synchronization result."""
+    offset: int
+    confidence: float
+    dists: list[float]
+    track_id: NotRequired[int]
+
+class PipelineResult(TypedDict):
+    """Complete pipeline result."""
+    video_path: str | Path
+    sync_results: list[SyncResult]
+    num_tracks: int
+    processing_time: float
+    error: NotRequired[str]
+
+# Model types
+if 'TypeAlias' in globals() and TypeAlias != type:
+    ModelState: TypeAlias = dict[str, torch.Tensor]
+    EmbeddingBatch: TypeAlias = torch.Tensor  # Shape: [batch_size, embedding_dim]
+else:
+    # Python 3.9 compatibility
+    ModelState = Dict[str, torch.Tensor]
+    EmbeddingBatch = torch.Tensor

{syncnet_python-0.1.0 → syncnet_python-0.1.1}/syncnet_python/__init__.py

@@ -4,7 +4,7 @@ This package provides a PyTorch implementation of SyncNet for detecting
 synchronization between audio and video in multimedia content.
 """
 
-__version__ = "0.1.0"
+__version__ = "0.1.1"
 
 # Import main components
 try:

{syncnet_python-0.1.0 → syncnet_python-0.1.1}/syncnet_python.egg-info/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.4
 Name: syncnet-python
-Version: 0.1.0
-Summary: SyncNet: Audio-visual synchronization detection using deep learning
+Version: 0.1.1
+Summary: SyncNet: Audio-visual synchronization detection using deep learning. Updated version of https://github.com/joonson/syncnet_python for modern Python versions.
 Author: SyncNet Python Contributors
 Maintainer: SyncNet Python Contributors
 License: MIT
@@ -49,6 +49,8 @@ Dynamic: license-file
 
 Audio-visual synchronization detection using deep learning.
 
+This is an updated version of the original [SyncNet implementation](https://github.com/joonson/syncnet_python) by Joon Son Chung, compatible with modern Python versions (3.9+).
+
 ## Overview
 
 SyncNet Python is a PyTorch implementation of the SyncNet model, which detects audio-visual synchronization in videos. It can identify lip-sync errors by analyzing the correspondence between mouth movements and spoken audio.
@@ -126,9 +128,13 @@ syncnet-python video.mp4 --device cpu
 - CUDA (optional but recommended)
 - FFmpeg
 
+## Credits
+
+This package is based on the original [SyncNet implementation](https://github.com/joonson/syncnet_python) by Joon Son Chung.
+
 ## Citation
 
-If you use this code in your research, please cite:
+If you use this code in your research, please cite the original paper:
 
 ```bibtex
 @inproceedings{chung2016out,

syncnet_python-0.1.0/syncnet/core/types.py

@@ -1,48 +0,0 @@
-"""Type definitions for SyncNet components."""
-
-from pathlib import Path
-from typing import TypeAlias, TypedDict, NotRequired
-import numpy as np
-import torch
-
-# Basic type aliases
-BBox: TypeAlias = tuple[float, float, float, float]
-Frame: TypeAlias = np.ndarray
-AudioData: TypeAlias = np.ndarray
-MFCCFeatures: TypeAlias = np.ndarray
-
-# Detection types
-class Detection(TypedDict):
-    """Face detection result."""
-    frame_idx: int
-    bbox: BBox
-    confidence: float
-    landmarks: NotRequired[np.ndarray]
-
-# Track types
-class Track(TypedDict):
-    """Face track information."""
-    frame: np.ndarray
-    bbox: np.ndarray
-    start_frame: int
-    end_frame: int
-
-# Result types
-class SyncResult(TypedDict):
-    """Synchronization result."""
-    offset: int
-    confidence: float
-    dists: list[float]
-    track_id: NotRequired[int]
-
-class PipelineResult(TypedDict):
-    """Complete pipeline result."""
-    video_path: str | Path
-    sync_results: list[SyncResult]
-    num_tracks: int
-    processing_time: float
-    error: NotRequired[str]
-
-# Model types
-ModelState: TypeAlias = dict[str, torch.Tensor]
-EmbeddingBatch: TypeAlias = torch.Tensor  # Shape: [batch_size, embedding_dim]