Systems and methods for signaling of downsampling filters for chroma from luma intra prediction mode
The various implementations described herein include methods and systems for decoding video. In one aspect, a method includes receiving a video stream having a sequence of frames. The sequence of frames includes one or more key frames. Each key frame has a respective downsampling filter type. The method includes, in accordance with a determination that a current frame corresponds to a first key frame, retrieving, from the video stream, a syntax element associated with a first downsampling filter type associated with the first key frame. The method includes a luma block of the current frame and a predefined set of frames immediately after the current frame using the first downsampling filter type to obtain downsampled frames. The method includes predicting a chroma block of the current frame based on the downsampled frames.
1. A method for video decoding, performed at a computing system having one or more processors and memory, the method comprising:
receiving a video stream having a sequence of frames, the sequence of frames including one or more key frames, and each key frame having a respective downsampling filter type;
in accordance with a determination that a current frame corresponds to a first key frame of the one or more key frames, wherein the one or more key frames are associated with N downsampling filter types, N being an integer;
retrieving, from the video stream, a syntax element associated with a first downsampling filter type associated with the first key frame, wherein the syntax element has a fixed length coding corresponding to M bits, M being an integer and satisfying a relationship N≤2 M ;
downsampling a luma block of the current frame and a predefined set of frames immediately after the current frame using the first downsampling filter type to obtain downsampled frames; and
predicting a chroma block of the current frame based on the downsampled frames.
2. The method of claim 1 , wherein the current frame includes a plurality of color components, the method further comprising:
downsampling a first color component of the plurality of color components of the current frame to obtain a first downsampled color component; and
predicting other color components of the plurality of color components based on the first downsampled color component.
3. The method of claim 1 , wherein the one or more key frames comprise an intra frame.
4. The method of claim 1 , wherein the one or more key frames correspond to every intra frame of the video stream.
5. The method of claim 1 , wherein the one or more key frames comprise an Instantaneous Decoder Refresh (IDR) frame.
6. The method of claim 1 , wherein the one or more key frames comprise one or more inter frames, each of the one or more inter frames is associated with a temporal layer ID lower than a predetermined threshold value.
7. The method of claim 1 , wherein the first downsampling filter type is a filter type corresponding to a Chroma from Luma (CfL) prediction mode for the video stream.
8. The method of claim 1 , wherein the syntax element includes a first attribute having an attribute value that indicates whether the first downsampling filter type is a 6-tap filter.
9. The method of claim 1 , wherein the syntax element is binarized using a Unary code.
10. The method of claim 1 , wherein the syntax element is binarized using a Truncated Unary code.
11. The method of claim 1 , wherein the syntax element is binarized using a Truncated Rice code.
12. The method of claim 1 , wherein the syntax element is binarized using am Exp-Golomb code.
13. The method of claim 1 , wherein:
the syntax element includes a first attribute having an attribute value that is determined based on whether the current frame corresponds to detected screen content.
14. The method of claim 13 , wherein:
when the current frame corresponds to detected screen content, the attribute value indicates whether co-located luma samples are used without applying any downsampling filter; and
when the current frame does not correspond to detected screen content, the attribute value indicates whether the first downsampling filter type is a 6-tap filter.
15. A method for video encoding performed at a computing system having one or more processors and memory, the method comprising:
receiving video data comprising a sequence of frames, the sequence of frames including one or more key frames, and each key frame having a respective downsampling filter type;
determining that a current frame corresponds to a first key frame of the one or more key frames, wherein the one or more key frames are associated with N downsampling filter types, N being an integer; and
encoding the current frame;
wherein a video bitstream comprises the encoded current frame and a syntax element indicating a first downsampling filter type for the first key frame, wherein the syntax element has a fixed length coding corresponding to M bits, M being an integer and satisfying a relationship N≤2 M .
16. The method of claim 15 , wherein the one or more key frames comprise an intra frame.
17. The method of claim 15 , wherein the one or more key frames correspond to every intra frame of the video bitstream.
18. The method of claim 15 , wherein the one or more key frames comprise an Instantaneous Decoder Refresh (IDR) frame.
19. The method of claim 15 , wherein the one or more key frames comprise one or more inter frames, each of the one or more inter frames is associated with a temporal layer ID lower than a predetermined threshold value.
20. A method of processing visual media data, the method comprising:
obtaining a source video sequence that comprises a plurality of frames; and
performing a conversion between the source video sequence and a video bitstream of visual media data according to a format rule,
wherein the video bitstream comprises a sequence of frames, the sequence of frames including one or more key frames, and each key frame having a respective downsampling filter type; and
wherein the format rule specifies that when it is determined that a current frame corresponds to a first key frame of the one or more key frames, and the one or more key frames are associated with N downsampling filter types, N being an integer:
a syntax element is to be retrieved from the video bitstream, the syntax element associated with a first downsampling filter type associated with the first key frame, wherein the syntax element has a fixed length coding corresponding to M bits, M being an integer and satisfying a relationship N≤2 M ;
a luma block of the current frame and a predefined set of frames are to be downsampled immediately after the current frame using the first downsampling filter type to obtain downsampled frames; and
a chroma block of the current frame is predicted based on the downsampled frames.