IP Library Granted Patent US 12,464,117
Granted Patent B2
US 12,464,117 · App. 18/438,202 · Granted Nov 4, 2025

Complexity reduction and bit-width control for bi-directional optical flow

Inventors: Xiaoyu Xiu (San Diego, CA); Yuwen He (San Diego, CA); Yan Ye (San Diego, CA); Jiancong Luo (Skillman, NJ)
Assignee: INTERDIGITAL VC HOLDINGS, INC.
H04N19/105H04N19/132H04N19/146H04N19/159H04N19/176H04N19/513
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,464,117
App. No.
18/438,202
Granted
Nov 4, 2025
Kind
B2
Abstract

Systems and methods are described for reducing the complexity of using bi-directional optical flow (BIO) in video coding. In some embodiments, bit-width reduction steps are introduced in the BIO motion refinement process to reduce the maximum bit-width used for BIO calculations. In some embodiments, simplified interpolation filters are used to generate predicted samples in an extended region around a current coding unit. In some embodiments, different interpolation filters are used for vertical versus horizontal interpolation. In some embodiments, BIO is disabled for coding units with small heights and/or for coding units that are predicted using a sub-block level inter prediction technique, such as advanced temporal motion vector prediction (ATMVP) or affine prediction.

Claims (97)

1 . A video decoding method comprising:

obtaining a prediction for a current block in a video with bi-directional optical flow;

decoding the current block based on the prediction;

wherein obtaining the prediction with bi-directional optical flow comprises:

obtaining a first prediction signal array from a first reference picture;

obtaining a first array of horizontal gradients including performing a right bit shift on two samples in the first prediction signal array and determining a difference between the two right-bit-shifted samples from the first prediction signal array;

obtaining a second prediction signal array from a second reference picture;

obtaining a second array of horizontal gradients including performing a right bit shift on two samples in the second prediction signal array and determining a difference between the two right-bit-shifted samples from the second prediction signal array;

wherein a size of the first prediction signal array and a size of the second prediction signal array is a size of the current block extended to one row on top and one row at bottom of a current block's boundary and to one column to a left and one column to a right of the current block's boundary;

obtaining a right bit shifted horizontal intermediate parameter array based on the first array of horizontal gradients and the second array of horizontal gradients,

wherein for a sample location of the current block that uses gradient values of samples that are outside the first and second arrays of horizontal gradients for determining a value of the right bit shifted horizontal intermediate parameter array, the gradient values of the samples that are outside the first and second arrays of horizontal gradients are set to gradient values of nearest neighbor samples inside the first and second arrays of horizontal gradients;

obtaining at least a horizontal motion refinement based at least in part on the horizontal first component intermediate parameter array; and

generating the prediction of the current block using at least the horizontal motion refinement.

2 . The method of claim 1 , wherein obtaining the right bit shifted horizontal intermediate parameter array based on the first array of horizontal gradients and the second array of horizontal gradients includes performing, for each sample in the horizontal intermediate parameter array, a right bit shift on a sum of (i) a corresponding sample in the first array of horizontal gradients and (ii) a corresponding sample in the second array of horizontal gradients.

3 . The method of claim 1 , wherein obtaining the prediction with bi-directional optical flow further comprises:

obtaining a signal-difference parameter array including performing a right bit shift on each of the first prediction signal array and the second prediction signal array and obtaining a difference between the shifted first prediction signal array and the shifted second prediction signal array, and

obtaining a signal-horizontal-gradient correlation parameter by summing components of an elementwise multiplication of the signal-difference parameter array with the horizontal intermediate parameter array;

wherein obtaining the horizontal motion refinement comprises bit-shifting the signal-horizontal-gradient correlation parameter to obtain the horizontal motion refinement.

4 . The method of claim 3 , wherein, for a sample location of the current block that uses prediction signal values that are outside the first prediction signal array and the second prediction signal array for calculating a value of the signal-difference parameter array, the prediction signal values of the samples that are outside the first prediction signal array and the second prediction signal array are set to prediction signal values of nearest neighbor samples inside the first prediction signal array and the second prediction signal array.

5 . The method of claim 1 , wherein obtaining the prediction with bi-directional optical flow further comprises:

obtaining a first array of vertical gradients including performing a right bit shift on two samples in the first prediction signal array and determining a difference between the two right-bit-shifted samples from the first prediction signal array;

obtaining a second array of vertical gradients including performing a right bit shift on two samples in the second prediction signal array and determining a difference between the two right-bit-shifted samples from the second prediction signal array;

obtaining a vertical intermediate parameter array by performing a right bit shift on a sum of the first array of vertical gradients and the second array of vertical gradients;

wherein, for a sample location of the current block that uses gradient values of samples that are outside the first and second arrays of vertical gradients for determining a value of the right bit shifted vertical intermediate parameter array, the gradient values of the samples that are outside the first and second arrays of vertical gradients are set to gradient values of nearest neighbor samples inside the first and second arrays of vertical gradients; and

obtaining a vertical motion refinement based at least in part on the vertical intermediate parameter array;

wherein the prediction of the current block is generated using the horizontal motion refinement and the vertical motion refinement.

6 . The method of claim 5 , wherein obtaining the prediction with bi-directional optical flow further comprises:

obtaining a cross-gradient correlation parameter by summing components of an elementwise multiplication of (i) the horizontal intermediate parameter array with (ii) the vertical intermediate parameter array;

wherein obtaining the vertical motion refinement comprises determining a product of (i) the horizontal motion refinement and (ii) the cross-gradient correlation parameter.

7 . The method of claim 5 , wherein generating the prediction comprises:

for each sample in the current block, calculating a bi-directional-optical-flow sample offset based on: (i) the first horizontal gradient array, (ii) the first vertical gradient array, (iii) the second horizontal gradient array, (iv) the second vertical gradient array, (v) the horizontal motion refinement, and (vi) the vertical motion refinement; and

for each sample in the current block, calculating a sum of at least the first prediction signal array, the second prediction signal array, and the bi-directional-optical-flow sample offset.

8 . A video decoder apparatus comprising one or more processors configured to perform at least:

obtaining a prediction for a current block in a video with bi-directional optical flow;

decoding the current block based on the prediction;

wherein obtaining the prediction with bi-directional optical flow comprises:

obtaining a first prediction signal array from a first reference picture;

obtaining a first array of horizontal gradients including performing a right bit shift on two samples in the first prediction signal array and determining a difference between the two right-bit-shifted samples from the first prediction signal array;

obtaining a second prediction signal array from a second reference picture;

obtaining a second array of horizontal gradients including performing a right bit shift on two samples in the second prediction signal array and determining a difference between the two right-bit-shifted samples from the second prediction signal array;

wherein a size of the first prediction signal array and a size of the second prediction signal array is a size of the current block extended to one row on top and one row at bottom of a current block's boundary and to one column to a left and one column to a right of the current block's boundary;

obtaining a right bit shifted horizontal intermediate parameter array based on the first array of horizontal gradients and the second array of horizontal gradients;

wherein for a sample location of the current block that uses gradient values of samples that are outside the first and second arrays of horizontal gradients for determining a value of the right bit shifted horizontal intermediate parameter array, the gradient values of the samples that are outside the first and second arrays of horizontal gradients are set to gradient values of nearest neighbor samples inside the first and second arrays of horizontal gradients;

obtaining at least a horizontal motion refinement based at least in part on the horizontal intermediate parameter array; and

generating the prediction of the current block using at least the horizontal motion refinement.

9 . The apparatus of claim 8 , wherein obtaining the right bit shifted horizontal intermediate parameter array based on the first array of horizontal gradients and the second array of horizontal gradients includes performing, for each sample in the horizontal intermediate parameter array, a right bit shift on a sum of (i) a corresponding sample in the first array of horizontal gradients and (ii) a corresponding sample in the second array of horizontal gradients.

10 . The apparatus of claim 8 , wherein obtaining the prediction with bi-directional optical flow further comprises:

obtaining a signal-difference parameter array including performing a right bit shift on each of the first prediction signal array and the second prediction signal array and obtaining a difference between the shifted first prediction signal array and the shifted second prediction signal array, and

obtaining a signal-horizontal-gradient correlation parameter by summing components of an elementwise multiplication of the signal-difference parameter array with the horizontal intermediate parameter array;

wherein obtaining the horizontal motion refinement comprises performing a left bit shift on the signal-horizontal-gradient correlation parameter to obtain the horizontal motion refinement.

11 . The apparatus of claim 10 , wherein obtaining the prediction with bi-directional optical flow comprises:

obtaining a first array of vertical gradients including performing a right bit shift on two samples in the first prediction signal array and determining a difference between the two right-bit-shifted samples from the first prediction signal array;

obtaining a second array of vertical gradients including performing a right bit shift on two samples in the second prediction signal array and determining a difference between the two right-bit-shifted samples from the second prediction signal array;

obtaining a vertical intermediate parameter array by performing a right bit shift on a sum of the first array of vertical gradients and the second array of vertical gradients;

obtaining a signal-vertical-gradient correlation parameter by summing components of an elementwise multiplication of the signal-difference parameter array with the vertical intermediate parameter array;

performing a left bit shift on the signal-vertical-gradient correlation parameter; and

obtaining a vertical motion refinement based at least in part on the vertical intermediate parameter array and the signal-vertical-gradient correlation parameter;

wherein the prediction of the current block is generated using the horizontal motion refinement and the vertical motion refinement.

12 . The apparatus of claim 11 , wherein the horizontal intermediate parameter array and the vertical intermediate parameter array are obtained using a same number of right bit shifts.

13 . The apparatus of claim 8 , wherein obtaining a prediction for a current block in a video with bi-directional optical flow is disabled for the current block responsive to a determination that the current block is coded in at least one of a subblock mode, an affine prediction mode, a merge mode with motion vector difference, the current block is not coded by a decoder-side motion vector derivation with bilateral matching, an inter and intra combined multi-hypothesis prediction, or a multi-hypothesis inter prediction prediction.

14 . The apparatus of claim 10 , wherein obtaining a prediction for a current block in a video with bi-directional optical flow is disabled for the current block if the current block has a height of 4 or for 4×8 size block.

15 . A video encoding method comprising:

obtaining a prediction for a current block in a video with bi-directional optical flow;

encoding the current block based on the prediction;

wherein obtaining the prediction with bi-directional optical flow comprises:

obtaining a first prediction signal array from a first reference picture;

obtaining a first array of horizontal gradients including performing a right bit shift on two samples in the first prediction signal array and determining a difference between the two right-bit-shifted samples from the first prediction signal array;

obtaining a second prediction signal array from a second reference picture;

obtaining a second array of horizontal gradients including performing a right bit shift on two samples in the second prediction signal array and determining a difference between the two right-bit-shifted samples from the second prediction signal array;

wherein a size of the first prediction signal array and a size of the second prediction signal array is a size of the current block extended to one row on top and one row at bottom of a current block's boundary and to one column to a left and one column to a right of the current block's boundary;

obtaining a right bit shifted horizontal intermediate parameter array based on the first array of horizontal gradients and the second array of horizontal first component gradients;

wherein for a sample location of the current block that uses gradient values of samples that are outside the first and second arrays of horizontal gradients for determining a value of the right bit shifted horizontal intermediate parameter array, the gradient values of the samples that are outside the first and second arrays of horizontal gradients are set to gradient values of nearest neighbor samples inside the first and second arrays of horizontal gradients;

obtaining at least a horizontal motion refinement based at least in part on the horizontal intermediate parameter array; and

generating the prediction of the current block using at least the horizontal motion refinement.

16 . The method of claim 15 , wherein obtaining the right bit shifted horizontal intermediate parameter array based on the first array of horizontal gradients and the second array of horizontal gradients includes performing, for each sample in the horizontal intermediate parameter array, a right bit shift on a sum of (i) a corresponding sample in the first array of horizontal gradients and (ii) a corresponding sample in the second array of horizontal gradients.

17 . The method of claim 15 , wherein obtaining the prediction with bi-directional optical flow further comprises:

obtaining a signal-difference parameter array including performing a right bit shift on each of the first prediction signal array and the second prediction signal array and obtaining a difference between the shifted first prediction signal array and the shifted second prediction signal array, and

obtaining a signal-horizontal-gradient correlation parameter by summing components of an elementwise multiplication of the signal-difference parameter array with the horizontal intermediate parameter array;

wherein obtaining the horizontal motion refinement comprises performing a left bit shift on the signal-horizontal-gradient correlation parameter to obtain the horizontal motion refinement.

18 . A video encoder apparatus comprising one or more processors configured to perform:

obtaining a prediction for a current block in a video with bi-directional optical flow;

encoding the current block based on the prediction;

wherein obtaining the prediction with bi-directional optical flow comprises:

obtaining a first prediction signal array from a first reference picture;

obtaining a first array of horizontal gradients including performing a right bit shift on two samples in the first prediction signal array and determining a difference between the two right-bit-shifted samples from the first prediction signal array;

obtaining a second prediction signal array from a second reference picture;

obtaining a second array of horizontal gradients including performing a right bit shift on two samples in the second prediction signal array and determining a difference between the two right-bit-shifted samples from the second prediction signal array;

wherein a size of the first prediction signal array and a size of the second prediction signal array is a size of the current block extended to one row on top and one row at bottom of a current block's boundary and to one column to a left and one column to a right of the current block's boundary;

obtaining a right bit shifted horizontal intermediate parameter array based on the first array of horizontal gradients and the second array of horizontal gradients;

wherein for a sample location of the current block that uses gradient values of samples that are outside the first and second arrays of horizontal gradients for determining a value of the right bit shifted horizontal intermediate parameter array, the gradient values of the samples that are outside the first and second arrays of horizontal gradients are set to gradient values of nearest neighbor samples inside the first and second arrays of horizontal gradients;

obtaining at least a horizontal motion refinement based at least in part on the horizontal intermediate parameter array; and

generating the prediction of the current block using at least the horizontal motion refinement.

19 . The apparatus of claim 18 , wherein obtaining the right bit shifted horizontal intermediate parameter array based on the first array of horizontal gradients and the second array of horizontal gradients includes performing, for each sample in the horizontal intermediate parameter array, a right bit shift on a sum of (i) a corresponding sample in the first array of horizontal gradients and (ii) a corresponding sample in the second array of horizontal gradients.

20 . The apparatus of claim 18 , wherein obtaining the prediction with bi-directional optical flow further comprises:

obtaining a signal-difference parameter array including performing a right bit shift on each of the first prediction signal array and the second prediction signal array and obtaining a difference between the shifted first prediction signal array and the shifted second prediction signal array, and

obtaining a signal-horizontal-gradient correlation parameter by summing components of an elementwise multiplication of the signal-difference parameter array with the horizontal intermediate parameter array;

wherein obtaining the horizontal motion refinement comprises performing a left bit shift on the signal-horizontal-gradient correlation parameter to obtain the horizontal motion refinement.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 19, 2024
From: XIU, XIAOYU; HE, YUWEN; YE, YAN; LUO, JIANCONG
To: VID SCALE, INC.
Reel/Frame 068035/0372 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 11, 2024
From: VID SCALE, INC.
To: INTERDIGITAL VC HOLDINGS, INC.
Reel/Frame 068284/0031 →
Continuity (8)
Continuation 17962928 · Oct 10, 2022
Continuation 17276448
Continuation 17962918 · Oct 10, 2022
Continuation 17276448
Provisional Application 62789331 · Jan 7, 2019
Provisional Application 62738655 · Sep 28, 2018
Provisional Application 62734763 · Sep 21, 2018
Related Publication 20240205389A1 · Jun 20, 2024
References Cited (61)
US 10931969B2 · Chen · 2021 [cited by applicant]
US 10986367B2 · Lee · 2021 [cited by examiner]
US 11153599B2 · Su · 2021 [cited by applicant]
US 11206419B2 · Lee · 2021 [cited by examiner]
US 11470308B2 · Xiu · 2022 [cited by applicant]
US 11470348B2 · Su · 2022 [cited by applicant]
US 11575933B2 · He · 2023 [cited by applicant]
US 20120170650A1 · Chong · 2012 [cited by applicant]
US 20130028530A1 · Drugeon · 2013 [cited by applicant]
US 20180098066A1 · Lee · 2018 [cited by applicant]
US 20180192072A1 · Chen · 2018 [cited by applicant]
US 20180241998A1 · Chen · 2018 [cited by applicant]
US 20180242004A1 · Park · 2018 [cited by applicant]
US 20180376166A1 · Chuang · 2018 [cited by applicant]
US 20190138889A1 · Jiang · 2019 [cited by applicant]
US 20190273938A1 · Abe · 2019 [cited by applicant]
US 20200128264A1 · Wang · 2020 [cited by applicant]
US 20200221122A1 · Ye · 2020 [cited by examiner]
US 20200280735A1 · Lim · 2020 [cited by applicant]
US 20210250592A1 · Xiu · 2021 [cited by applicant]
US 20230050213A1 · Xiu · 2023 [cited by applicant]
CN 107646195 · 2018 [cited by applicant]
JP H06168239 · 1994 [cited by applicant]
KR 101547742B1 · 2015 [cited by applicant]
KR 1020180043787 · 2018 [cited by applicant]
TW 201826791 · 2018 [cited by applicant]
WO 2011129084 · 2011 [cited by applicant]
WO 2018084339 · 2018 [cited by applicant]
WO 2018113658 · 2018 [cited by applicant]
WO 2018166357 · 2018 [cited by applicant]
WO 2019238008 · 2019 [cited by applicant]
WO 2020035054 · 2020 [cited by applicant]
ITU-T Recommendation H.264 and ISO/IEC/MPEG-4 part 10, “Advanced Video Coding for Generic Audiovisual Services”, Series H: Audiovisual and Multimedia Systems, Infrastructure of audiovisual services—Coding of moving vide… [cited by applicant]
Bross, et. al., “High Efficiency Video Coding (HEVC) Text Specification Draft 10 (for FDIS and Last Call)”. Joint Collaborative Team on Video Coding (JCT-VC), Document No. JCTVC-L1003, Jan. 2013, 310 pages. [cited by applicant]
Alshina, E., et. al., “Known Tools Performance Investigation for Next Generation Video Coding”. ITU-Telecommunications Standardization Sector, SG16/Q6, Video Coding Experts Group (VCEG), Jun. 2015, VCEG-AZ05, 7 pages. [cited by applicant]
Alshin, Alexander, et. al., “Bi-directional Optical Flow for Future Video Codec”. IEEE Data Compression Conference (DCC), (2016), pp. 83-90. [cited by applicant]
Wikipedia, “Karatsuba Algorithm”. Wikipedia web article, available at: https://web.archive.org/web/20170303213652/https://en.wikipedia.org/wiki/Karatsuba algorithm, Mar. 2017 (5 pages). [cited by applicant]
Xiu, Xiaoyu, et. al., “CE9-Related: Complexity Reduction and Bit-Width Control for Bi-Directional Optical Flow (BIO)”. Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-L0256, Oct. … [cited by applicant]
Xiu, Xiaoyu, et. al., “CE9-Related: A simplified design of bi-direction optical flow (BIO)”. JVET Meeting, JVET-L0591, 12th meeting Oct. 2018 (9 pages). [cited by applicant]
International Search Report and Written Opinion of the International Searching Authority for PCT/US2019/051560 mailed Dec. 4, 2019, 13 pages. [cited by applicant]
Segall, Andrew, et. al. “Joint Call For Proposals on Video Compression With Capability Beyond HEVC”. Joint Video Exploration Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-H1002-v6, Oct. 18-24, 2017… [cited by applicant]
Bross, Benjamin, et. al., “Versatile Video Coding (Draft 2)”. Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-K1001-v6, Jul. 2018, 140 pages. [cited by applicant]
Jeong, Seungsoo, et. al., “CE4 Ultimate Motion Vector Expression (Test 4.5.4)”. Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-L0054, Oct. 2018, 6 pages. [cited by applicant]
Jeong, Seungsoo, et. al., “Proposed WD for CE4 Ultimate motion vector expression (Test 4.5.4)”. Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-L0054_WD, Oct. 2018, 12 pages. [cited by applicant]
Chiang, Man-Shu, et. al., “CE10.1.1: Multi-Hypothesis Prediction for Improving AMVP Mode, Skip or Merge Mode, and Intra Mode”. Joint Video Experts Team (JVET)of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-L0100… [cited by applicant]
Chiang, Man-Shu, et. al., “CE10.1.1: Multi-Hypothesis Prediction for Improving AMVP Mode, Skip or Merge Mode, and Intra Mode”. Joint Video Experts Team (JVET)of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-L0100… [cited by applicant]
Winken, Martin, et. al., “CE10-Related: Multi-Hypothesis Inter Prediction with simplified AMVP Process”. Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-L0679-v2, Oct. 2018, 4 pag… [cited by applicant]
Winken, Martin, et. al., “CE10-Related: Multi-Hypothesis Inter Prediction with simplified AMVP Process”. Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-L0679-v2, Power Point Pres… [cited by applicant]
Esenlik, Semih, et. al., “Simplified DMVR for Inclusion in VVC”. Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-L0670, Oct. 2018, 5 pages. [cited by applicant]
International Preliminary Report on Patentability for PCT/US2019/051560 issued on Mar. 23, 2021, 9 pages. [cited by applicant]
SMPTE Standard “VC-1 Compressed Video Bitstream Format and Decoding Process”. The society of motion picture and television engineers, SMPTE 421M, 2006, (493 pages). [cited by applicant]
Chen , Jianle, et. al., “Coding Tools Investigation for Next Generation Video Coding”. ITU-Telecommunication Standardization Sector, Study Group 16, Contribution 806, COM16-C806, Jan. 2015, pp. 1-7. [cited by applicant]
Karczewicz, M., et. al., “Report of AHG1 on Coding Efficiency Improvements”. ITU-Telecommunications Standardization Sector, Video Coding Experts Group (VCEG), SG16/Q6, VCEG-AZ01, Jun. 2015, 2 pages. [cited by applicant]
Ohm, Jens-Rainer., et. al., “Report of AHG on Future Video Coding Standardization Challenges”. International Organization for Standardization, Coding of Moving Pictures and Audio, ISO/IEC JTC1/SC29/WG11 MPEG2014/M36782,… [cited by applicant]
Jingya, L., “CE9-related: Simplification of BIO”. Joint Video Exploration Team (JVET) of ITU-T SG 16 WP 3 and ISO/ IEC JTC 1/SC 29/WG 11, JVET-L0123-v2, 12th meeting, Macao, CN, Oct. 2018 (9 pages). [cited by applicant]
Xiu, X., et al., “CE9.5.3: Bi-directional optical flow (BIO) simplification”. Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-K0344-v1, 11th meeting, Ljubljana, SI, Jul. 2018 (2 p… [cited by applicant]
Chen, C.Y., et al., “CE9.5.2: Bio with simplified gradient calculation, adaptive BIO granularity, and applying BIO to chroma components” Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11,… [cited by applicant]
Hung, C.H., et al., “CE9: Bio gradient calculation improvement (test 9.5)” Joint Video Exploration Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, JVET-K0119, 11th meeting, Ljubljana, SI, Jul. 2017 (4 pag… [cited by applicant]
Lai, C. Y et al., “CE9-related: BIO Simplification”. Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/ IEC JTC 1/SC 29/WG 11, JVET-L0099-v3, 12th meeting, Macao, CN, Oct. 2018 (6 pages). [cited by applicant]
Xiu et al., “Description of SDR, HDR and 360° Video Coding Technology Proposal by InterDigital Communications and Dolby Laboratories”, JVET-J0015-V1, InterDigital Communications, Inc., Dolby Laboratories, Inc., Joint Vi… [cited by applicant]
Xiu, X., et al., “CE9-related: A simplified bi-directional optical flow (BIO) design based on the combination of CE9.5.2 test 1 and CE9.5.3” Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG… [cited by applicant]