Unified neural network in-loop filter

1. A method implemented by a video coding apparatus, comprising:

applying a neural network (NN) filter to an unfiltered sample of a video unit to generate a filtered sample, wherein the NN filter is based on an NN filter model generated using a quality-level indicator (QI) input; and

converting between a video media file and a bitstream based on the filtered sample that was generated.

2. The method of claim 1 , wherein the QI input comprises a quantization parameter (QP) of the video unit.

3. The method of claim 1 , wherein the QI input comprises a quantization parameter (QP) of a sequence that includes the video unit.

4. The method of claim 1 , wherein the QI input comprises a quantization parameter (QP) of a picture or slice that includes the video unit.

5. The method of claim 1 , wherein the QI input comprises an output of a quantization process applied to the video unit, wherein the quantization process is based on a quantization parameter (QP) of the video unit, a QP of a sequence that includes the video unit, or a QP of a picture or slice that includes the video unit.

6. The method of claim 1 , further comprising spanning the quality-level indicator into a two-dimensional array having a same size as the video unit, wherein the two-dimensional array is an additional input plane to the NN filter.

7. The method of claim 6 , wherein the video unit is a coding tree unit (CTU) or a coding unit (CU).

8. The method of claim 1 , wherein the QI input is fed into a plurality of fully-connected layers of the NN filter to expand the QI input into a one-dimensional vector, and wherein elements of the one-dimensional vector are channel-wise scaling factors used to recalibrate features maps of the NN filter model.

9. The method of claim 8 , wherein the feature maps of the NN filter model are given by f∈R N×W×H , where N represents the channel numbers, W represents the channel width, and H represents the channel height, wherein the one-dimensional vector provided by the fully-connected layers is represented by S∈R N , and wherein the recalibrated feature maps are generated according to F i,j,k =f i,j,k ×S i , 1≤i≤N, 1≤j≤W, 1≤k≤H, where F represents the recalibrated feature maps.

10. The method of claim 1 , further comprising signaling the quality-level indicator to a decoder using one selected from the group consisting of: a sequence parameter set (SPS), a picture parameter set (PPS), a picture header, a slice header, a coding tree unit (CTU), and a coding unit (CU).

11. The method of claim 1 , wherein the video unit comprises first and second color components, wherein the QI input indicates a quality of the first color component, and wherein the NN filter is applied to the first and second color components to generate the filtered sample.

12. The method of claim 1 , wherein the video unit comprises first and second color components, wherein the QI input comprises a first QI that indicates a quality of the first color component and a second QI that indicates a quality of the second color component, and wherein the first QI is applied to the first color component and the second QI is applied to the second color component to generate the filtered sample.

13. The method of claim 1 , wherein the video unit comprises first and second color components, wherein the QI input indicates a quality of the first color component, and wherein the NN filter is applied only to the first color component to generate the filtered sample.

14. The method of claim 1 , wherein the QI indicates a quality of a region or picture that includes the video unit.

15. The method of claim 1 , wherein the QI indicates a quality of a region or picture that does not include the video unit.

16. The method of claim 15 , wherein the video unit refers to the region or picture.

17. The method of claim 1 , wherein the conversion comprises generating the bitstream according to the video media file.

18. The method of claim 1 , wherein the conversion comprises parsing the bitstream to obtain the video media file.

19. An apparatus for coding video data comprising a processor and a non-transitory memory with instructions thereon, wherein the instructions upon execution by the processor cause the processor to:

apply a neural network (NN) filter to an unfiltered sample of a video unit to generate a filtered sample, wherein the NN filter is based on an NN filter model generated using a quality-level indicator (QI) input; and

convert between a video media file and a bitstream based on the filtered sample that was generated.

20. A non-transitory computer readable medium storing a bitstream of a video that is generated by a method performed by a video processing apparatus, wherein the method comprises:

applying a neural network (NN) filter to an unfiltered sample of a video unit to generate a filtered sample, wherein the NN filter is based on an NN filter model generated using a quality-level indicator (QI) input; and

generating the bitstream based on the filtered sample that was generated.