# US 12,225,184 B2 · Virtual memory access bandwidth verification (VMBV) in video coding > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,225,184 B2 - Document ID: US12225184B2 - Status: Granted - Kind: B2 - Application number: 18/214,017 - Filed: 2023-06-26 - Published: 2023-10-19 - Granted: 2025-02-11 - Earliest claimed: 2011-10-04 - Assignees: Texas Instruments Incorporated - Applicants: Texas Instruments Incorporated - Inventors: Minhua Zhou - Web page: https://patentgrep.com/patent/US12225184B2 ## CPC - H04N 19/00: Methods or arrangements for coding, decoding, compressing or decompressing digital video signals - H04N 19/433: characterised by techniques for memory access - H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION ## Abstract A method is provided that includes determining a target picture virtual memory access (VMA) bandwidth rate, wherein the target picture VMA bandwidth rate indicates a maximum VMA bandwidth rate for motion compensation of a picture, and verifying the target picture VMA bandwidth rate for a compressed video bit stream. ## Claims ### Claim 1 A method comprising: encoding, by a processor, a first coding unit (CU) in a picture; determining, by the processor, a consumed virtual memory access (VMA) bandwidth for the first CU; limiting, by the processor, a maximum number of reference pictures for motion estimation for a second CU in the picture based on the consumed VMA bandwidth for the first CU; and encoding, by the processor, the second CU using the limited maximum number of reference pictures for motion estimation. ### Claim 2 (depends on 1) The method of claim 1, further comprising: adapting a target CU VMA bandwidth for the second CU based on the consumed VMA bandwidth for the first CU; and determining that the adapted target CU VMA bandwidth for the second CU is less than a threshold level, wherein limiting the maximum number of reference pictures for motion estimation for the second CU is in response to determining that the adapted target CU VMA bandwidth for the second CU is less than the threshold level. ### Claim 3 (depends on 2) The method of claim 2, wherein determining that the adapted target CU VMA bandwidth for the second CU is less than the threshold level comprises determining that the adapted target CU VMA bandwidth for the second CU is less than zero. ### Claim 4 (depends on 2) The method of claim 2, wherein the threshold level is a first threshold level, wherein the method further comprises determining that the adapted target CU VMA bandwidth for the second CU is greater than a second threshold level, and wherein limiting the maximum number of reference pictures for motion estimation for the second CU comprises setting the maximum number of reference pictures to two in response to determining that the adapted target CU VMA bandwidth for the second CU is less than the first threshold level and greater than the second threshold level. ### Claim 5 (depends on 2) The method of claim 2, wherein the threshold level is a first threshold level, wherein the method further comprises determining that the adapted target CU VMA bandwidth for the second CU is less than a second threshold level, and wherein limiting the maximum number of reference pictures for motion estimation for the second CU comprises setting the maximum number of reference pictures to one in response to determining that the adapted target CU VMA bandwidth for the second CU is less than the first threshold level and less than the second threshold level. ### Claim 6 (depends on 1) The method of claim 1, wherein the first and second CUs are first and second largest CUs (LCUs), wherein the method further comprises disabling 8×8 CUs for the second LCU based on the consumed VMA bandwidth for the first LCU, and wherein encoding the second LCU comprises encoding the second LCU without 8×8 CUs. ### Claim 7 (depends on 1) The method of claim 1, wherein the first CU is a first largest CU (LCU) comprising one or more inter-predicted CUs and one or more intra-predicted CUs, and wherein determining the consumed VMA bandwidth for the first LCU is based on the one or more inter-predicted CUs and not based on the one or more intra-predicted CUs. ### Claim 8 (depends on 7) The method of claim 7, wherein determining the consumed VMA bandwidth for the first LCU is based on a sum of bandwidths of the one or more inter-predicted CUs. ### Claim 9 (depends on 1) The method of claim 1, further comprising: determining a target picture VMA bandwidth rate for the picture; determining a target CU VMA bandwidth for the first CU based on the target picture VMA bandwidth rate; and adapting the target CU VMA bandwidth for the second CU based on the consumed VMA bandwidth for the first CU, wherein limiting the maximum number of reference pictures for motion estimation for the second CU is based on the adapted target CU VMA bandwidth for the second CU. ### Claim 10 A system comprising: a processor configured to: receive a video sequence including a picture; and transform the video sequence into coding units (CUs); and an encoder configured to: encode a first CU in the picture; determine a consumed virtual memory access (VMA) bandwidth for the first CU; limit a maximum number of reference pictures for motion estimation for a second CU in the picture based on the consumed VMA bandwidth for the first CU; and encode the second CU using the limited maximum number of reference pictures for motion estimation. ### Claim 11 (depends on 10) The system of claim 10, wherein the encoder is further configured to: adapt a target CU VMA bandwidth for the second CU based on the consumed VMA bandwidth for the first CU; and determine that the adapted target CU VMA bandwidth for the second CU is less than a threshold level, and wherein the encoder is configured to limit the maximum number of reference pictures for motion estimation for the second CU in response to determining that the adapted target CU VMA bandwidth for the second CU is less than the threshold level. ### Claim 12 (depends on 11) The system of claim 11, wherein to determine that the adapted target CU VMA bandwidth for the second CU is less than the threshold level, the encoder is configured to determine that the adapted target CU VMA bandwidth for the second CU is less than zero. ### Claim 13 (depends on 11) The system of claim 11, wherein the threshold level is a first threshold level, wherein the encoder is further configured to determine that the adapted target CU VMA bandwidth for the second CU is greater than a second threshold level, and wherein to limit the maximum number of reference pictures for motion estimation for the second CU, the encoder is configured to set the maximum number of reference pictures to two in response to determining that the adapted target CU VMA bandwidth for the second CU is less than the first threshold level and greater than the second threshold level. ### Claim 14 (depends on 11) The system of claim 11, wherein the threshold level is a first threshold level, wherein the encoder is further configured to determine that the adapted target CU VMA bandwidth for the second CU is less than a second threshold level, and wherein to limit the maximum number of reference pictures for motion estimation for the second CU, the encoder is configured to set the maximum number of reference pictures to one in response to determining that the adapted target CU VMA bandwidth for the second CU is less than the first threshold level and less than the second threshold level. ### Claim 15 (depends on 10) The system of claim 10, wherein the first and second CUs are first and second largest CUs (LCUs), wherein the encoder is further configured to disable 8×8 CUs for the second LCU based on the consumed VMA bandwidth for the first LCU, and wherein to encode the second LCU, the encoder is configured to encode the second LCU without 8×8 CUs. ### Claim 16 (depends on 10) The system of claim 10, wherein the first CU is a first largest CU (LCU) comprising one or more inter-predicted CUs and one or more intra-predicted CUs, and wherein the encoder is configured to determine the consumed VMA bandwidth for the first LCU based on the one or more inter-predicted CUs and not based on the one or more intra-predicted CUs. ### Claim 17 (depends on 16) The system of claim 16, wherein the encoder is configured to determine the consumed VMA bandwidth for the first LCU based on a sum of bandwidths of the one or more inter-predicted CUs. ### Claim 18 (depends on 10) The system of claim 10, wherein the encoder is further configured to: determine a target picture VMA bandwidth rate for the picture; determine a target CU VMA bandwidth for the first CU based on the target picture VMA bandwidth rate; and adapt the target CU VMA bandwidth for the second CU based on the consumed VMA bandwidth for the first CU, wherein the encoder is configured to limit the maximum number of reference pictures for motion estimation for the second CU based on the adapted target CU VMA bandwidth for the second CU. ### Claim 19 A method comprising: encoding, by a processor, a first largest coding unit (LCU) in a picture; determining, by the processor, a consumed virtual memory access (VMA) bandwidth for the first LCU; and encoding, by the processor, a second LCU in the picture without 8×8 CUs based on the consumed VMA bandwidth for the first LCU. ### Claim 20 (depends on 19) The method of claim 19, further comprising: adapting a target LCU VMA bandwidth for the second LCU based on the consumed VMA bandwidth for the first LCU; and determining that the adapted target LCU VMA bandwidth for the second LCU is less than a threshold level; and disabling 8×8 CUs for the second LCU in response to determining that the adapted target LCU VMA bandwidth for the second LCU is less than the threshold level.