Nokia Technologies Oy

Nokia Corporation

Core Wireless Licensing S.a.r.l.

Microsoft Technology Licensing, LLC

Emre Baris Aksu

Miska Matias Hannuksela

Igor Danilo Diego Curcio

Sujeet Shyamsundar Mate

Francesco Cricri

Kashyap Kammachi Sreedhar

Hamed Rezazadegan Tavakoli

Igor D Curcio

Lukasz Kondrad

Francesco Cricrì

Yu You

David Leon

Honglei Zhang

Jani Lainema

Sebastian Schwarz

Saba Ahsan

Ari Hourunranta

Miikka Lundan

Peter O Fasogbon

Ru-Shang Wang

Kashyap Kammachi-Sreedhar

Ye-Kui Wang

Mikko Johannes Honkala

Ville-Veikko Mattila

Lauri Ilola

Lauri Aleksi Ilola

Viktor Varsa

Xingyang Ni

Nam Le

Rod Walsh

Caglar Aytekin

Harsh Mehta

Marta Karczewicz

Ossi Kalevo

Kimmo Tapio Roimela

Antti Hallapuro

Iraj Saniee

Vinod Kumar Malamal Vadakital

Lixin Fan

Juha-Pekka Hippeläinen

Jonne Juhani Mäkinen

Miika Sakari Tupala

Joachim Wabnig

Keith D Miller

Mika Pesonen

Alireza Aminlou

Ville Timonen

Johannes Pystynen

Ramin Ghaznavi Youvalari

Alireza Zare

Nannan Zou

Igor D D Curcio

Johannes Rajala

Olli Kilpelainen

Jani Kattelus

Igor Danilo Curcio

Peter Oluwanisola Fasogbon

Miika Tupala

Miska Hannuksela

The embodiments relate to a method for encoding two or more tensors. The method comprises processing the two or more tensors having respective dimensions so that the dimensions of said two or more sensors have the same number (); identifying which axis of each individual tensor is swappable to result in concatenable tensors around an axis of concatenation (); reshaping the tensors so that the dimensions are modified based on the swapped axis (); concatenating the tensors around the axis of concatenation to result in concatenated tensor (); compressing the concatenated tensor (); generating syntax structures for carrying concatenation and axis swapping information (); and generating a bitstream by combining the syntax structures and the compressed concatenated tensor (). The embodiments also relate to a method for decoding, and to apparatuses for implementing the methods.

Method, an apparatus and a computer program product for neural network compression

A method comprising obtaining a 360-degree video content from a video source; projecting the 360-degree video content onto a 2D image plane; dividing the projected 360-degree video content into a plurality of regions, wherein the regions are partly overlapping and each region covers a region of the 360-degree video content suitable for a viewport presentation; receiving a request for a viewport orientation of the 360-degree video content from a client; and providing the client with an viewport presentation of the region corresponding to the requested viewport orientation.

Method, an apparatus and a computer program product for video conferencing

There is disclosed an apparatus and a method for spatial computing service session description for volumetric XR conversation. In accordance an embodiment the method comprises receiving volumetric video from a plurality of sources, the volumetric video data comprising color images and depth images representing at least a part of a scene; determining which of the plurality of sources belong to a same group; selecting a group identifier for the group; associating the volumetric video with the group identifier indicative of from which group the volumetric video data is received; associating the volumetric video with a source identifier indicative of from which source of the group the volumetric video data is received; associating the color image of the volumetric video with a color data identifier; and associating the depth image of the volumetric video with a depth data identifier.

Method, an apparatus and a computer program product for spatial computing service session description for volumetric extended reality conversation

In example embodiments, an apparatus, a method, and a computer program product are provided. The apparatus comprises at least one processor; and at least one non-transitory memory including computer program code; wherein the at least one memory and the computer program code are configured to, with the at least one processor, cause the apparatus at least to perform determine a processing order of building blocks to encode or decode a media item; and determine a number of processing steps required to encode or decode the media item; wherein the processing order of building blocks and the number of processing steps are determined based on a content of the media item by using a guided probability model based on a neural network.

Guided probability model for compressed representation of neural networks

An apparatus includes at least one processor; and at least one non-transitory memory including computer program code; wherein the at least one memory and the computer program code are configured to, with the at least one processor, cause the apparatus at least to: estimate an importance of parameters of a neural network based on a graph diffusion process over at least one layer of the neural network; determine the parameters of the neural network that are suitable for pruning or sparsification; remove neurons of the neural network to prune or sparsify the neural network; and provide at least one syntax element for signaling the pruned or sparsified neural network over a communication channel, wherein the at least one syntax element comprises at least one neural network representation syntax element.

Graph diffusion for structured pruning of neural networks

A method is provided for defining a metadata box of a neural network representation (NNR) item data, wherein the NNR item data comprises an NNR bitstream; and defining an association between the NNR item data and an NNR configuration by using a configuration item property, wherein the NNR configuration item property comprises information about stored NNR item data. Corresponding apparatuses and computer program products are also provided.

Carriage and signaling of neural network representations

A method, apparatus and computer program product are provided to signaling multiple media data types in a container file format, for example, ISOBMFF, in an efficient manner. In the context of a signaling method, media type definitions overloaded in an instance a plurality of media type definitions is present in compatible media type list, wherein a media type definition is associated with a media type. The method then indicates, using the compatible media type list, presence of a plurality of media types in a metadata or multimedia box. Corresponding apparatuses and computer program products are also provided.

Method and apparatus for signaling multiple media data types

There are disclosed various methods, apparatuses and computer program products for video encoding and decoding. In some embodiments the method for video encoding comprises obtaining compressed volumetric video data representing a three-dimensional scene or object (); capsulating the compressed volumetric video data into a data structure (); obtaining data of a two-dimensional projection of at least a part of the three-dimensional scene as seen from a certain viewport (); and including the data of the two-dimensional projection into the data structure ().

Apparatus, a method and a computer program for omnidirectional video

Example embodiments provide a system for training a data coding pipeline including a feature extractor neural network, an encoder neural network, and a decoder neural network configured to reconstruct input data based on encoded features. A plurality of losses corresponding to different tasks may be determined for the coding pipeline. Tasks may be performed based on an output of the coding pipeline. A weight update may be determined for at least a subset of the coding pipeline based on the plurality of losses. The weight update may be configured to reduce a number of iterations for fine-tuning the coding pipeline for one of the tasks. This enables faster adaptation of the coding pipeline for one of the tasks after deployment of the coding pipeline. Apparatuses, methods, and computer programs are disclosed. Apparatuses, methods, and computer programs are disclosed.

Training a data coding system comprising a feature extractor neural network

The embodiments relate to a method comprising establishing a three-dimensional conversational interaction with one or more receivers; generating a pointcloud relating to a user and capturing audio from one or more audio source; generating conversational scene description comprising at least a first dynamic object describing a virtual space for the three-dimensional conversational interaction, wherein the first dynamic object refers to one or more objects specific to the three-dimensional conversational interaction, wherein said one or more objects comprises at least data relating to transformable pointcloud; audio obtained from said one or more audio source and input obtained from one or more connected devices controlling at least the pointcloud, wherein said objects are linked to each other for seamless manipulation; applying the conversational scene description into a metadata, and transmitting the metadata with the respective audio in realtime to said one or more receivers.

Method, an apparatus and a computer program product for video encoding and video decoding

A method, apparatus and computer program product are provided to include information within a container that also includes a video bitstream as to whether individual image items corresponding to the video frames of the video bitstream bitstreams are self-decodable or, alternatively, are dependent upon one or more other image items. In an instance in which a respective image item is dependent upon one or more other image items, the method, apparatus and computer program product also include dependence information within the container identifying the other image item(s)upon which the decodability of the respective image item is dependent. As such, the method, apparatus and computer program product permit decoding relationships to be defined in the container between a predictively coded frame and other image item(s)upon which the predictively coded frame is dependent, thereby facilitating the decoding of the frame and, in turn, the video bitstream.

Method and apparatus for storing and signaling predictively coded image items

A method comprising: authoring a plurality of sets of media tracks comprising at least a first set of media tracks and a second set of media tracks into a media file format, wherein a subset of tracks of the first set comprises alternate data for each other and a subset of tracks of the second set comprises alternate data for each other; and including, in or along a bitstream comprising a media file including or inferring said media tracks, an indication that said subset of tracks of the first set are alternatives to each other and said subset of tracks of the second set are alternatives to each other upon playback of the media tracks.

Apparatus, a method and a computer program for video coding and decoding

A method, an apparatus, and a computer program product are provided. An example method includes defining an enhancement message comprising at least one of the following: an identifying number for identifying a post-processing filter; a mode identity (idc) field used of indicating association of a post-processing filter with the identifying number; a flag for specifying the enhancement message being used for a current layer; and the payload byte comprising a bitstream; and using the enhancement message for at least one of specifying a neural network that is used as a post-processing filter or cancelling a use of a previous post-processing filter with the same identifying number.

High-level syntax for signaling neural networks within a media bitstream

The embodiments relate to a method for encoding and a decoding, and apparatuses for the same. The method for encoding comprises receiving a block of a video frame for encoding (); making a decision on whether or not a learning-based model is to be applied as a processing step for encoding the block (); applying the learning-based model for said input block according to the decision, where the learning-based model has been selectively fine-tuned according to information relating to activation of the learning-based model of previously-decoded blocks (); encoding a signal corresponding to the decision on usage of the learning-based model into a bitstream (); and encoding the block into a bitstream with an information whether the block is to be used for finetuning ().

The embodiments relate to a method for writing, in a container file, two or more subpicture tracks; writing, in the container file, a base track, which is intended to be resolved into a video bitstream; indicating, in the base track, a layout of subpictures; writing, in the container file, a sample group description entry indicative of a first subpicture track or a group of subpicture tracks for each subpicture position in the layout of subpictures, wherein the first subpicture track includes the subpicture sequence for the respective subpicture position and wherein any track among the group of subpicture tracks includes a valid subpicture sequence for the respective subpicture position; and indicating in the container file, samples of the base track for which the sample group description entry is intended to be used for reconstructing the video bitstream. The embodiments also relate to a method for parsing, as well as technical equipment for implementing the method for writing and the method for parsing.

Growing community of inventors

Tampere, Finland

Emre Baris Aksu