US 11,889,066 B2
Intra-prediction mode concept for block-wise picture coding
Jonathan Pfaff, Berlin (DE); Philipp Helle, Berlin (DE); Dominique Maniry, Berlin (DE); Thomas Wiegard, Berlin (DE); Wojciech Samek, Berlin (DE); Stephan Kaltenstadler, Berlin (DE); Heiko Schwarz, Berlin (DE); Detlev Marpe, Berlin (DE); Mischa Siekmann, Berlin (DE); and Martin Winken, Berlin (DE)
Assigned to Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V., Munich (DE)
Filed by Fraunhofer-Gesellschaft Zur Förderung derangewandten Forschung e.V., Munich (DE)
Filed on Jun. 13, 2022, as Appl. No. 17/839,459.
Application 17/839,459 is a continuation of application No. 16/845,715, filed on Apr. 10, 2020, granted, now 11,363,259.
Application 16/845,715 is a continuation of application No. PCT/EP2018/077609, filed on Oct. 10, 2018.
Claims priority of application No. 17196402 (EP), filed on Oct. 13, 2017.
Prior Publication US 2022/0321881 A1, Oct. 6, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. H04N 19/11 (2014.01); H04N 19/109 (2014.01); H04N 19/176 (2014.01); H04N 19/61 (2014.01)
CPC H04N 19/109 (2014.11) [H04N 19/176 (2014.11); H04N 19/61 (2014.11)] 18 Claims
OG exemplary drawing
 
1. An apparatus for designing a neural network or a sequence of one or more linear functions in which each linear function is followed by a respective non-linear function for assisting in selecting among a set of intra-prediction modes for block-based picture coding, the apparatus comprising:
a processor; and
a memory storing instructions that, when executed by the processor, cause the processor to:
predict, using each of the set of intra-prediction modes, a first picture test block from a first set of neighboring samples neighboring the first picture test block to acquire, for each of the set of intra-prediction modes, a prediction signal for the first picture test block;
apply the first set of neighboring samples onto one of the neural network or the sequence of one or more linear functions in which each linear function is followed by a respective non-linear function to acquire, for each of the set of intra-prediction modes, a probability value indicative of a probability of a respective intra-prediction mode;
determine, for each of the set of intra-prediction modes, a cost estimate for coding costs related to prediction error coding and mode signalization using the prediction signal acquired for the respective intra-prediction mode;
update parameters of the neural network or the sequence of one or more linear functions to reduce a coding cost function having a first addend forming a residual rate estimate depending on the prediction signal acquired for the intra-prediction mode of lowest coding cost estimate, and a second addend forming a mode signaling side information rate estimate depending on the prediction signal and the probability value acquired for the intra-prediction mode of lowest coding cost estimate; and
use the updated parameters to:
predict, using each of the set of intra-prediction modes, a second picture test block from a second set of neighboring samples neighboring the second picture test block to acquire, for each of the set of intra-prediction modes, a prediction signal for the second picture test block; and
apply the second set of neighboring samples onto one of the neural network or the sequence of one or more linear functions in which each linear
function is followed by a respective non-linear function to acquire, for each of the set of intra-prediction modes, a probability value indicative of a probability of the respective intra-prediction mode.