DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Priority
Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55.
Information Disclosure Statement
The information disclosure statement (IDS) was submitted on 06/07/2024. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-15 are rejected under 35 U.S.C. 103 as being unpatentable over Pogorelik et al. (“Pogorelik”) [PGPUB 2021/0406652] in view of Zhao [PGPUB 2024/0022616] (the cited subject matter is supported by provisional application 63/397,281).
Regarding claim 1, the Pogorelik reference discloses a method performed by a user equipment (UE) for an artificial intelligence/machine learning (AI/ML) media service in a wireless communication system [ie. client (“UE”) with face recognition service (“AI/ML media service”); Pogorelik; figures 4 and 5B-5C; paragraph 0026, 0029-0030, and 0047], the method comprising:
receiving, from a network server providing the AI/ML media service, service access information [ie. server inquires or discovers security requirements (“service access information”) from client (“receiving” at UE); Pogorelik; fig 4 and 6; para 0063, 0070, and 0089];
obtaining information on client AI media inferencing capabilities and functions [ie. obtain capabilities and functions to be sent to server; Pogorelik; fig 6; para 0063; 0077, 0081, and 0126];
negotiating with the network server for splitting an AI media inference processing, based on the received service access information and the obtained information on client AI media inferencing capabilities and functions [Pogorelik; para 0054-0055, 0063-0064, and 0087]; and
receiving, from the network server, either intermediate data or inference output data by AI model split inferencing [server sends messages (“intermediate data”) to receiving client (“UE”), in order for the client to produce the final inference output; Pogorelik; fig 9A; para 0101-0102].
The Pogorelik reference discloses a client receiving discovery inquiry for access security requirements and establishing a trust from a network server providing the AI/ML media service [Pogorelik; fig 4 and 6; para 0063, 0070, and 0088-0089] but does not specifically state “service access information including at least one of information for media session handling and information for media streaming access”.
However, in the same field of endeavor, the Zhao reference discloses receiving, from a network server providing the AI/ML media service, service access information including at least one of information for media session handling and information for media streaming access [ie. “5GMS defined media-streaming architecture for both uplink and downlink streaming. A 5GMS-aware application is enabled to utilize the MS interface for media session handling and the M4 interface for streaming transport handling”; Zhao; fig 1; para 0021, 0024-0025, and 0123]. The Pogorelik and Zhao references are analogous art, since they have similar problem solving area in data channel management for media streaming. It would have been obvious to a person of ordinary skill in the art, before the effective filling date of the claimed invention, to combine the teaching of media session handling and access, taught by Zhao, into the system, taught by Pogorelik. The motivation for doing so would have been “for immersive RTC signaling and streaming based on SGMS existing interface” [Zhao; para 0025].
Regarding claim 2, the combination of Pogorelik-Zhao further discloses AI model data related to a structure of an AI model for the AI/ML media service includes a UE AI model subset and a network AI model subset [Pogorelik; para 0063 and 0128] [Zhao; para 0123].
Regarding claim 3, the combination of Pogorelik-Zhao further discloses UE AI model data corresponding to the UE AI model subset is provided to the UE by the network server [Pogorelik; para 0063 and 0128] [Zhao; para 0123].
Regarding claim 4, the combination of Pogorelik-Zhao further discloses outputting inference output data based on the UE AI model data and the received intermediate data [Pogorelik; para 0095 and 0101-0103] [Zhao; para 0123].
Regarding claim 5, the combination of Pogorelik-Zhao further discloses outputting intermediate data by performing AI model split inferencing based on the UE AI model data; and transmitting, to the network server, the outputted intermediate data [ie. server receives back propagating gradients (“intermediate data”), by AI client-side model split, and transmitted by the client; Pogorelik; para 0052 and 0095] [Zhao; para 0176-0177].
Regarding claims 6-10, the apparatus of claims 6-10 perform the similar steps as the method of claims 1-6. The combination of Pogorelik-Zhao teaches the method of claims 1-6, as referenced above. The additional limitations of an “network server”, a “transceiver”, and a “processor” are rejected with the citation of paragraphs 0034-0036 of Pogorelik. Therefore, claims 6-10 are rejected using the same art and rationale set forth above in the rejection of claims 1-6, by the teachings of Pogorelik-Zhao.
Regarding claim 11, the Pogorelik reference discloses a for an artificial intelligence/machine learning (AI/ML) media service in a wireless communication system, the network server comprising: a transceiver; and a processor configured to: [Pogorelik; figures 4 and 5B-5C; paragraph 0026, 0029-0030, 0034-0036, and 0047]:
transmit, to a user equipment (UE) via the transceiver, service access information [ie. server inquires or discovers (“transmits”) security requirements (“service access information”) from client and establishes trust; Pogorelik; fig 4 and 6; para 0063, 0070, 0085, and 0089]
negotiate with the UE for splitting an AI media inference processing, based on the transmitted service access information [Pogorelik; para 0054-0055, 0063-0064, and 0087], and
transmit, to the UE via the transceiver, either intermediate data or inference output data by AI model split inferencing [server sends messages (“intermediate data”) to receiving client (“UE”), in order for the client to produce the final inference output; Pogorelik; fig 9A; para 0101-0102].
The Pogorelik reference discloses a client receiving discovery inquiry for access security requirements and establishing a trust from a network server providing the AI/ML media service [Pogorelik; fig 4 and 6; para 0063, 0070, and 0088-0089] but does not specifically state “service access information including at least one of information for media session handling and information for media streaming access”.
However, in the same field of endeavor, the Zhao reference discloses service access information including at least one of information for media session handling and information for media streaming access [ie. “SGMS defined media-streaming architecture for both uplink and downlink streaming. A SGMS-aware application is enabled to utilize the MS interface for media session handling and the M4 interface for streaming transport handling”; Zhao; fig 1 and 10; para 0021, 0024-0025, and 0114]. The Pogorelik and Zhao references are analogous art, since they have similar problem solving area in data channel management for media streaming. It would have been obvious to a person of ordinary skill in the art, before the effective filling date of the claimed invention, to combine the teaching of media session handling and access, taught by Zhao, into the system, taught by Pogorelik. The motivation for doing so would have been “for immersive RTC signaling and streaming based on SGMS existing interface” [Zhao; para 0025].
Regarding claim 12, the combination of Pogorelik-Zhao further discloses AI model data related to a structure of an AI model for the AI/ML media service includes a UE AI model subset and a network AI model subset [Pogorelik; para 0063 and 0128] [Zhao; fig 10; para 0114].
Regarding claim 13, the combination of Pogorelik-Zhao further discloses provide UE AI model data corresponding to the UE AI model subset to the UE [Pogorelik; para 0063 and 0128] [Zhao; fig 10; para 0114].
Regarding claim 14, the combination of Pogorelik-Zhao further discloses output the intermediate data by performing the AI model split inferencing based on the network AI model subset [Pogorelik; para 0095 and 0101-0103] [Zhao; fig 10; para 0114].
Regarding claim 15, the combination of Pogorelik-Zhao further discloses receive, via the transceiver from the UE, intermediate data based on the UE AI model data, and output the inference output data by performing the AI model split inferencing based on the received intermediate data and the network AI model subset [ie. server receives back propagating gradients (“intermediate data”), by AI client-side model split, and transmitted by the client; Pogorelik; para 0052 and 0095] [Zhao; para 0176-0177].
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Yip et al [PGPUB 2025/0056256 and PGPUB 2024/0381127] are the same assignee that describe similar inventions of AI split modeling.
Guan et al. [PGPUB 2023/0328492] describes multiple UE with inference result sharing.
Kashyap et al. [PGPUB 2024/0292198] describes AI split modeling with multiple clients within a proximity of each other.
Kovacs et al. [PGPUB 2024/0196187] describes discovering UE capabilities with sidelink.
Khoa et al. [NPL “splitDyn”] describes training a split AI model that is on multiple clients.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JASON D CARDONE whose telephone number is (571)272-3933. The examiner can normally be reached Mon-Fri. 8am-4pmEST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Umar Cheema can be reached at 571-270-3037. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/JASON D CARDONE/Primary Examiner, Art Unit 2458
Prosecution Insights