Description:AS ATTACHED , Claims:I/We Claim:
1. Data stream (45) having a representation of a neural network (10) encoded
thereinto, wherein the data stream (45) is structured into one or more
5 individually accessible portions (200), each individually accessible portion
representing a corresponding neural network layer (210, 30) of the neural
network, wherein the data stream (45) further comprises, for a
predetermined neural network layer, a neural network layer type parameter
(130) indicating a neural network layer type of the predetermined neural
10 network layer of the neural network.
2. Apparatus for encoding a representation of a neural network (10) into a data
stream (45), so that the data stream (45) is structured into one or more
individually accessible portions (200), each individually accessible portion
15 representing a corresponding neural network layer (210, 30) of the neural
network, wherein the apparatus is configured to provide the data stream (45)
with, for a predetermined neural network layer, a neural network layer type
parameter (130) indicating a neural network layer type of the predetermined
neural network layer of the neural network.
20
3. Apparatus for decoding a representation of a neural network (10) from a data
stream (45), wherein the data stream (45) is structured into one or more
individually accessible portions (200), each portion representing a
corresponding neural network layer (210, 30) of the neural network, wherein
25 the apparatus is configured to decode from the data stream (45), for a
predetermined neural network layer (210, 30), a neural network layer type
parameter (130) indicating a neural network layer type of the predetermined
neural network layer of the neural network.
4. Apparatus of claim 4, wherein the neural network layer type parameter (130)
discriminates, at least, between a fully-connected and a convolutional layer
type.
5 5. Apparatus of claim 3 or claim 4, wherein the data stream (45), is structured
into individually accessible portions (200), each individually accessible
portion representing a corresponding neural network portion of the neural
network, and
wherein the apparatus is configured to decode, from the data stream (45),
10 for each of one or more predetermined individually accessible portions
(200), a pointer (220, 244) pointing to a beginning of each individually
accessible portion.
6. Apparatus of claim 5, wherein each individually accessible portion
15 represents
a corresponding neural network layer (210) of the neural network or
a neural network portion (43, 44, 240) of a neural network layer (210) of the
neural network.
20 7. Apparatus of any of claims 3 to 6, wherein the apparatus is configured to
decode a representation of a neural network (10) from the data stream (45),
wherein the data stream (45) is structured into one or more individually
accessible portions (200), each individually accessible portion representing
a corresponding neural network layer (210, 30) of the neural network, and
25 wherein the data stream (45) is, within a predetermined portion, further
structured into individually accessible sub-portions (43, 44, 240), each sub-
portion (43, 44, 240) representing a corresponding neural network portion
of the respective neural network layer (210, 30) of the neural network,
wherein the apparatus is configured to decode from the data stream (45), for
30 each of one or more predetermined individually accessible sub-portions (43,
44, 240)
a start code (242) at which the respective predetermined individually
accessible sub-portion begins, and/or
a pointer (244) pointing to a beginning of the respective predetermined
individually accessible sub-portion, and/or
5 a data stream length parameter indicating a data stream length (246) of the
respective predetermined individually accessible sub-portion for skipping
the respective predetermined individually accessible sub-portion in parsing
the data stream (45).
10 8. Apparatus of claim 7, wherein the apparatus is configured to decode, from
the data stream (45), the representation of the neural network using context-
adaptive arithmetic decoding and using context initialization at a start of
each individually accessible portion and each individually accessible sub-
portion.
15
9. Apparatus of any previous claim 3 to 8, wherein the apparatus is configured
to decode a representation of a neural network (10) from a data stream (45),
wherein the data stream (45) is structured into individually accessible
portions (200), each portion representing a corresponding neural network
20 portion of the neural network, wherein the apparatus is configured to decode
from the data stream (45), for each of one or more predetermined
individually accessible portions, an identification parameter (310) for
identifying the respective predetermined individually accessible portion.
25 10. Apparatus of claim 9, wherein the identification parameter (310) is related
to the respective predetermined individually accessible portion via a hash
function or error detection code or error correction code.
11. Apparatus of claim 9 or claim 10, wherein the apparatus is configured to
30 decode, from the data stream (45), a higher-level identification parameter
(310) for identifying a collection of more than one predetermined
individually accessible portion.
12. Apparatus of claim 11, wherein the higher-level identification parameter
5 (310) is related to the identification parameters (310) of the more than one
predetermined individually accessible portion via a hash function or error
detection code or error correction code.
13. Apparatus of any previous claim 3 to 12, wherein the apparatus is configured
10 to decode a representation of a neural network (10) from a data stream (45),
wherein the data stream (45) is structured into individually accessible
portions (200), each portion representing a corresponding neural network
portion of the neural network, wherein the apparatus is configured to decode
from the data stream (45), for each of one or more predetermined
15 individually accessible portions a supplemental data (350) for
supplementing the representation of the neural network.
14. Apparatus of claim 13, wherein the data stream (45) indicates the
supplemental data (350) as being dispensable for inference based on the
20 neural network.
15. Apparatus of claim 13 or claim 14, wherein the apparatus is configured to
decode the supplemental data (350) for supplementing the representation of
the neural network for the one or more predetermined individually
25 accessible portions (200) from further individually accessible portions,
wherein the data stream (45) comprises for each of the one or more
predetermined individually accessible portions a corresponding further
predetermined individually accessible portion relating to the neural network
portion to which the respective predetermined individually accessible
30 portion corresponds.
16. Apparatus of any previous claim 13 to 15, wherein the supplemental data
(350) relates to
relevance scores of neural network parameters (32), and/or
5 perturbation robustness of neural network parameters (32).
17. Apparatus of any previous claim 3 to 16, for decoding a representation of a
neural network (10) from a data stream (45), wherein the apparatus is
configured to decode from the data stream (45) hierarchical control data
10 (400) structured into a sequence (410) of control data portions (420),
wherein the control data portions provide information on the neural network
at increasing details along the sequence of control data portions.
18. Apparatus of claim 17, wherein at least some of the control data portions
15 (420) provide information on the neural network which is partially
redundant.
19. Apparatus of claim 17 or claim 18, wherein a first control data portion
provides the information on the neural network by way of indicating a
20 default neural network type implying default settings and a second control
data portion comprises a parameter to indicate each of the default settings.
20. Apparatus for performing an inference using a neural network, comprising
25 an apparatus for decoding a data stream (45) according to any of claims 3 to
19, so as to derive from the data stream (45) the neural network, and
a processor configured to perform the inference based on the neural network.
30 21. Method for encoding a representation of a neural network into a data stream,
so that the data stream is structured into one or more individually accessible
portions, each individually accessible portion representing a corresponding
neural network layer of the neural network, wherein the method comprises
providing the data stream with, for a predetermined neural network layer, a
neural network layer type parameter indicating a neural network layer type
5 of the predetermined neural network layer of the neural network.
22. Method for decoding a representation of a neural network from a data
stream, wherein the data stream is structured into one or more individually
accessible portions, each portion representing a corresponding neural
10 network layer of the neural network, wherein the method comprises
decoding from the data stream, for a predetermined neural network layer, a
neural network layer type parameter indicating a neural network layer type
of the predetermined neural network layer of the neural network.
15 23. Computer program for, when executed by a computer, causing the computer
to perform the method of claim 21 or claim 22.