Meeting: TSGS4_135_India | Agenda Item: 9.7
4 documents found
[FS_AIF_MED] HDR Images
This Change Request proposes the addition of High Dynamic Range (HDR) imaging support to the AI for Media feasibility study. The CR introduces a new scenario for HDR images with supporting references and framework for evaluation.
The CR adds seven new normative/informative references to support HDR imaging:
Android Ultra HDR Image Format documentation [2]
ISO/IEC standards:
ISO 21496-1: Gain map metadata for dynamic range conversion [7]
3GPP and GSMA specifications:
Standards Alignment: The proposal builds upon existing ISO/IEC and 3GPP work, ensuring compatibility with established HDR imaging standards
Ecosystem Support: References both major mobile platforms (iOS and Android) demonstrating industry readiness
Messaging Integration: Leverages existing 3GPP TS 26.143 and GSMA RCS specifications for messaging applications
Evaluation Framework: Establishes placeholder structure for both objective and subjective evaluation methodologies (content pending)
This is an initial framework CR with several sections marked for future completion, particularly the evaluation criteria and methodology sections.
[FS_AIF_MED] Live Images
This change request proposes the addition of a new scenario to the FS_AIF_MED study item covering Live Images functionality.
The document introduces Live Images as a feature that combines: - A standard photograph - A short video component capturing movement
Key Characteristics: - Records moments in the immediate vicinity of the shutter press - Implemented across various consumer product ecosystems - References existing implementations from major platforms (Apple's Live Photos and Android's Motion Photos)
The contribution notes that: - Similar features in commercial products are already supported by the Image Sequence feature specified in ISO/IEC 23008-12 [3] - This provides a foundation for standardization work
The document establishes placeholders for: - Evaluation criteria and metrics (5.x.3) - marked as Editor's Note - Evaluation methodology (5.x.4) - marked as Editor's Note - Objective performance evaluation (5.x.4.1) - Subjective performance evaluation (5.x.4.2)
The CR adds three normative/informative references: 1. Apple Inc. developer documentation on Live Photos 2. Android Developers documentation on Motion Photo format 3. ISO/IEC 23008-12 (Image File Format specification)
This appears to be an initial text proposal establishing the framework for Live Images evaluation within the FS_AIF_MED study. The evaluation sections are incomplete and marked for future development.
[FS_AIF_MED] Stereo images
This CR proposes adding a new scenario to the FS_AIF_MED study focusing on the usage of stereoscopic images.
The contribution introduces the use case for immersive stereoscopic image formats:
The contribution references existing technologies and specifications:
The contribution includes placeholder sections for: - Evaluation criteria and metrics - Evaluation methodology (both objective and subjective performance evaluation)
These sections are marked with Editor's Notes indicating they are to be completed in future revisions.
The CR adds six normative/informative references: 1. Canon RF5.2mm F2.8 L Dual Fisheye documentation 2. Meta Quest immersive media formats documentation 3. Apple spatial photos and videos documentation 4. ISO/IEC 23008-12:2025 (HEIF) 5. ISO/IEC 23008-2:2025 (HEVC) 6. 3GPP TS 26.265 (Video Capabilities and Operation Points)
This appears to be an initial text proposal establishing the framework for studying stereoscopic images within the AI for Media feasibility study. The evaluation sections are placeholders awaiting further development and agreement.
[FS_AIF_MED] TR skeleton
This document presents a Technical Report (TR) skeleton for the Feasibility Study on AI/ML for Media (FS_AIF_MED). The contribution is submitted by Apple Inc. to SA4 Meeting #135 (India, 9-13 February 2026).
The document proposes a skeleton structure for the Technical Report that will capture the outcomes of the feasibility study on Artificial Intelligence and Machine Learning applications for media services. The TR skeleton serves as the foundational framework that will organize and document the study's findings, use cases, requirements, and conclusions.
The source requests SA4 to: - Agree to the TR skeleton as the basis for documenting the FS_AIF_MED study outcomes