route chunk read through frame_store (#1198)

* vision integration

* Deleted read_video_chunk + read_static_video_chunk

facefusion/processors/modules/age_modifier/core.py

@@ -23,7 +23,7 @@ from facefusion.processors.types import ProcessorOutputs
 from facefusion.program_helper import find_argument_group
 from facefusion.thread_helper import thread_semaphore
 from facefusion.types import ApplyStateItem, Args, DownloadScope, Face, InferencePool, ModelOptions, ModelSet, ProcessMode, VisionFrame
-from facefusion.vision import match_frame_color, read_static_image, read_static_video_chunk, read_static_video_frame
+from facefusion.vision import match_frame_color, read_static_image, read_static_video_frame
 
 
 @lru_cache()
@@ -166,7 +166,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	video_manager.clear_video_pool()
 
 	if state_manager.get_item('video_memory_strategy') in [ 'strict', 'moderate' ]:

facefusion/processors/modules/background_remover/core.py

@@ -22,7 +22,7 @@ from facefusion.program_helper import find_argument_group
 from facefusion.sanitizer import sanitize_int_range
 from facefusion.thread_helper import thread_semaphore
 from facefusion.types import ApplyStateItem, Args, DownloadScope, InferencePool, InferenceProvider, Mask, ModelOptions, ModelSet, ProcessMode, VisionFrame
-from facefusion.vision import read_static_image, read_static_video_chunk, read_static_video_frame
+from facefusion.vision import read_static_image, read_static_video_frame
 
 
 @lru_cache()
@@ -539,7 +539,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	video_manager.clear_video_pool()
 
 	if state_manager.get_item('video_memory_strategy') in [ 'strict', 'moderate' ]:

facefusion/processors/modules/deep_swapper/core.py

@@ -23,7 +23,7 @@ from facefusion.processors.types import ProcessorOutputs
 from facefusion.program_helper import find_argument_group
 from facefusion.thread_helper import thread_semaphore
 from facefusion.types import ApplyStateItem, Args, DownloadScope, Face, InferencePool, Mask, ModelOptions, ModelSet, ProcessMode, VisionFrame
-from facefusion.vision import conditional_match_frame_color, read_static_image, read_static_video_chunk, read_static_video_frame
+from facefusion.vision import conditional_match_frame_color, read_static_image, read_static_video_frame
 
 
 @lru_cache()
@@ -320,7 +320,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	video_manager.clear_video_pool()
 
 	if state_manager.get_item('video_memory_strategy') in [ 'strict', 'moderate' ]:

facefusion/processors/modules/expression_restorer/core.py

@@ -23,7 +23,7 @@ from facefusion.processors.types import LivePortraitExpression, LivePortraitFeat
 from facefusion.program_helper import find_argument_group
 from facefusion.thread_helper import conditional_thread_semaphore, thread_semaphore
 from facefusion.types import ApplyStateItem, Args, DownloadScope, Face, InferencePool, ModelOptions, ModelSet, ProcessMode, VisionFrame
-from facefusion.vision import read_static_image, read_static_video_chunk, read_static_video_frame
+from facefusion.vision import read_static_image, read_static_video_frame
 
 
 @lru_cache()
@@ -146,7 +146,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	video_manager.clear_video_pool()
 
 	if state_manager.get_item('video_memory_strategy') in [ 'strict', 'moderate' ]:

facefusion/processors/modules/face_debugger/core.py

@@ -19,7 +19,7 @@ from facefusion.processors.modules.face_debugger.types import FaceDebuggerInputs
 from facefusion.processors.types import ProcessorOutputs
 from facefusion.program_helper import find_argument_group
 from facefusion.types import ApplyStateItem, Args, Face, InferencePool, ProcessMode, VisionFrame
-from facefusion.vision import read_static_image, read_static_video_chunk, read_static_video_frame
+from facefusion.vision import read_static_image, read_static_video_frame
 
 
 def get_inference_pool() -> InferencePool:
@@ -68,7 +68,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	video_manager.clear_video_pool()
 
 	if state_manager.get_item('video_memory_strategy') == 'strict':

facefusion/processors/modules/face_editor/core.py

@@ -23,7 +23,7 @@ from facefusion.processors.types import LivePortraitExpression, LivePortraitFeat
 from facefusion.program_helper import find_argument_group
 from facefusion.thread_helper import conditional_thread_semaphore, thread_semaphore
 from facefusion.types import ApplyStateItem, Args, DownloadScope, Face, FaceLandmark68, InferencePool, ModelOptions, ModelSet, ProcessMode, VisionFrame
-from facefusion.vision import read_static_image, read_static_video_chunk, read_static_video_frame
+from facefusion.vision import read_static_image, read_static_video_frame
 
 
 @lru_cache()
@@ -197,7 +197,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	video_manager.clear_video_pool()
 
 	if state_manager.get_item('video_memory_strategy') in [ 'strict', 'moderate' ]:

facefusion/processors/modules/face_enhancer/core.py

@@ -21,7 +21,7 @@ from facefusion.processors.types import ProcessorOutputs
 from facefusion.program_helper import find_argument_group
 from facefusion.thread_helper import thread_semaphore
 from facefusion.types import ApplyStateItem, Args, DownloadScope, Face, InferencePool, ModelOptions, ModelSet, ProcessMode, VisionFrame
-from facefusion.vision import blend_frame, read_static_image, read_static_video_chunk, read_static_video_frame
+from facefusion.vision import blend_frame, read_static_image, read_static_video_frame
 
 
 @lru_cache()
@@ -337,7 +337,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	video_manager.clear_video_pool()
 
 	if state_manager.get_item('video_memory_strategy') in [ 'strict', 'moderate' ]:

facefusion/processors/modules/face_swapper/core.py

@@ -26,7 +26,7 @@ from facefusion.processors.types import ProcessorOutputs
 from facefusion.program_helper import find_argument_group
 from facefusion.thread_helper import conditional_thread_semaphore
 from facefusion.types import ApplyStateItem, Args, DownloadScope, Embedding, Face, InferencePool, InferenceProvider, ModelOptions, ModelSet, ProcessMode, VisionFrame
-from facefusion.vision import read_static_image, read_static_images, read_static_video_chunk, read_static_video_frame, unpack_resolution
+from facefusion.vision import read_static_image, read_static_images, read_static_video_frame, unpack_resolution
 
 
 @lru_cache()
@@ -588,7 +588,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	video_manager.clear_video_pool()
 
 	if state_manager.get_item('video_memory_strategy') in [ 'strict', 'moderate' ]:

facefusion/processors/modules/frame_colorizer/core.py

@@ -20,7 +20,7 @@ from facefusion.processors.types import ProcessorOutputs
 from facefusion.program_helper import find_argument_group
 from facefusion.thread_helper import thread_semaphore
 from facefusion.types import ApplyStateItem, Args, DownloadScope, InferencePool, InferenceProvider, ModelOptions, ModelSet, ProcessMode, VisionFrame
-from facefusion.vision import blend_frame, read_static_image, read_static_video_chunk, read_static_video_frame, unpack_resolution
+from facefusion.vision import blend_frame, read_static_image, read_static_video_frame, unpack_resolution
 
 
 @lru_cache()
@@ -230,7 +230,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	video_manager.clear_video_pool()
 
 	if state_manager.get_item('video_memory_strategy') in [ 'strict', 'moderate' ]:

facefusion/processors/modules/frame_enhancer/core.py

@@ -20,7 +20,7 @@ from facefusion.processors.types import ProcessorOutputs
 from facefusion.program_helper import find_argument_group
 from facefusion.thread_helper import conditional_thread_semaphore
 from facefusion.types import ApplyStateItem, Args, DownloadScope, InferencePool, InferenceProvider, ModelOptions, ModelSet, ProcessMode, VisionFrame
-from facefusion.vision import blend_frame, create_tile_frames, merge_tile_frames, read_static_image, read_static_video_chunk, read_static_video_frame
+from facefusion.vision import blend_frame, create_tile_frames, merge_tile_frames, read_static_image, read_static_video_frame
 
 
 @lru_cache()
@@ -623,7 +623,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	video_manager.clear_video_pool()
 
 	if state_manager.get_item('video_memory_strategy') in [ 'strict', 'moderate' ]:

facefusion/processors/modules/lip_syncer/core.py

@@ -23,7 +23,7 @@ from facefusion.processors.types import ProcessorOutputs
 from facefusion.program_helper import find_argument_group
 from facefusion.thread_helper import conditional_thread_semaphore
 from facefusion.types import ApplyStateItem, Args, AudioFrame, DownloadScope, Face, InferencePool, ModelOptions, ModelSet, ProcessMode, VisionFrame
-from facefusion.vision import read_static_image, read_static_video_chunk, read_static_video_frame
+from facefusion.vision import read_static_image, read_static_video_frame
 
 
 @lru_cache()
@@ -169,7 +169,6 @@ def pre_process(mode : ProcessMode) -> bool:
 def post_process() -> None:
 	read_static_image.cache_clear()
 	read_static_video_frame.cache_clear()
-	read_static_video_chunk.cache_clear()
 	read_static_voice.cache_clear()
 	video_manager.clear_video_pool()
 

facefusion/vision.py

@@ -10,7 +10,7 @@ from facefusion import ffprobe, video_manager
 from facefusion.common_helper import is_windows
 from facefusion.filesystem import get_file_extension, is_image, is_video
 from facefusion.thread_helper import thread_lock, thread_semaphore
-from facefusion.types import ColorMode, Duration, Fps, Mask, Orientation, Resolution, Scale, VisionFrame, VisionFrameSet
+from facefusion.types import ColorMode, Duration, Fps, Mask, Orientation, Resolution, Scale, VisionFrame
 
 
 def read_static_images(image_paths : List[str], color_mode : ColorMode = 'rgb') -> List[VisionFrame]:
@@ -90,37 +90,6 @@ def read_video_frame(video_path : str, frame_number : int = 0) -> Optional[Visio
 	return None
 
 
-@lru_cache(maxsize = 2)
-def read_static_video_chunk(video_path : str, chunk_number : int, chunk_size : int) -> VisionFrameSet:
-	return read_video_chunk(video_path, chunk_number, chunk_size)
-
-
-#todo: needs review - [decoding] [critical: medium] chunk end clamped by the ffprobe frame_total estimate
-def read_video_chunk(video_path : str, chunk_number : int, chunk_size : int) -> VisionFrameSet:
-	video_frame_chunk = {}
-
-	if is_video(video_path) and chunk_number > -1:
-		video_reader = video_manager.get_reader(video_path)
-
-		if video_reader:
-			video_frame_position = chunk_number * chunk_size
-			video_frame_end = video_frame_position + chunk_size
-
-			if video_reader.get('metadata').get('frame_total') > 0:
-				video_frame_end = min(video_frame_end, video_reader.get('metadata').get('frame_total'))
-
-			with thread_semaphore():
-				video_manager.conditional_set_video_reader_position(video_reader, video_frame_position)
-
-				for frame_number in range(video_frame_position, video_frame_end):
-					vision_frame = video_manager.read_video_reader_frame(video_reader)
-
-					if numpy.any(vision_frame):
-						video_frame_chunk[frame_number] = vision_frame
-
-	return video_frame_chunk
-
-
 #todo: needs review - [decoding] [critical: high] window read replaces the chunk cache, out of range frames fall back to empty vision frames
 #todo: question to restore the chunk_size approach over the window read
 def select_video_frames(video_path : str, frame_number : int = 0, frame_offset : int = 2) -> List[VisionFrame]:

tests/test_face_tracker.py

@@ -7,7 +7,7 @@ from facefusion.download import conditional_download
 from facefusion.face_creator import get_many_faces, get_one_face
 from facefusion.face_store import clear_faces
 from facefusion.face_tracker import create_face_tracks, select_face_track, track_faces
-from facefusion.vision import read_static_video_chunk, read_static_video_frame
+from facefusion.vision import read_static_video_frame, select_video_frames
 from .helper import get_test_example_file, get_test_examples_directory
 
 
@@ -47,8 +47,7 @@ def before_each() -> None:
 
 def test_track_faces() -> None:
 	target_path = get_test_example_file('target-240p.mp4')
-	video_frame_chunk = read_static_video_chunk(target_path, 0, 7)
-	target_vision_frames = [ video_frame_chunk.get(frame_number) for frame_number in sorted(video_frame_chunk) ]
+	target_vision_frames = select_video_frames(target_path, 3, 3)
 	empty_vision_frame = numpy.zeros_like(get_first(target_vision_frames))
 
 	target_vision_frames[2] = empty_vision_frame
@@ -58,7 +57,7 @@ def test_track_faces() -> None:
 
 	assert len(track_faces(target_vision_frames, 0.3)) == 1
 
-	target_vision_frames = [ video_frame_chunk.get(frame_number) for frame_number in sorted(video_frame_chunk)[:5] ]
+	target_vision_frames = select_video_frames(target_path, 3, 3)[:5]
 	target_vision_frames[0] = empty_vision_frame
 	target_vision_frames[1] = empty_vision_frame
 	target_vision_frames[2] = empty_vision_frame

tests/test_vision.py

@@ -6,7 +6,7 @@ import pytest
 from facefusion import ffmpeg, ffmpeg_builder, process_manager
 from facefusion.common_helper import is_linux
 from facefusion.download import conditional_download
-from facefusion.vision import calculate_histogram_difference, count_trim_frame_total, count_video_frame_total, detect_image_resolution, detect_video_duration, detect_video_fps, detect_video_resolution, match_frame_color, normalize_resolution, pack_resolution, predict_video_frame_total, read_image, read_video_chunk, read_video_frame, restrict_image_resolution, restrict_trim_frame, restrict_video_fps, restrict_video_resolution, scale_resolution, select_video_frames, unpack_resolution, write_image
+from facefusion.vision import calculate_histogram_difference, count_trim_frame_total, count_video_frame_total, detect_image_resolution, detect_video_duration, detect_video_fps, detect_video_resolution, match_frame_color, normalize_resolution, pack_resolution, predict_video_frame_total, read_image, read_video_frame, restrict_image_resolution, restrict_trim_frame, restrict_video_fps, restrict_video_resolution, scale_resolution, select_video_frames, unpack_resolution, write_image
 from .helper import get_test_example_file, get_test_examples_directory, get_test_output_file, prepare_test_output_directory
 
 
@@ -138,11 +138,6 @@ def test_read_video_frame() -> None:
 	assert read_video_frame('invalid') is None
 
 
-def test_read_video_chunk() -> None:
-	assert len(read_video_chunk(get_test_example_file('target-240p-25fps.mp4'), 1, 40)) == 40
-	assert read_video_chunk('invalid', 1, 40) == {}
-
-
 def test_select_video_frames() -> None:
 	assert len(select_video_frames(get_test_example_file('target-240p-25fps.mp4'), 50, 5)) == 11
 	assert len(select_video_frames(get_test_example_file('target-240p-25fps.mp4'), 1, 5)) == 11