LRN - 1 vs 13#
Next section compares an older to a newer version of the same operator after both definition are converted into markdown text. Green means an addition to the newer version, red means a deletion. Anything else is unchanged.
- LRN1 → LRN13 +0 -1
LRN1 → LRN13
RENAMED
@@ -1 +1 @@
|
|
1
1
|
Local Response Normalization proposed in the [AlexNet paper](https://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks.pdf).
|
2
2
|
It normalizes over local input regions.
|
3
3
|
The local region is defined across the channels. For an element X[n, c, d1, ..., dk] in a tensor
|
4
4
|
of shape (N x C x D1 x D2, ..., Dk), its region is
|
5
5
|
{X[n, i, d1, ..., dk] | max(0, c - floor((size - 1) / 2)) <= i <= min(C - 1, c + ceil((size - 1) / 2))}.
|
6
6
|
square_sum[n, c, d1, ..., dk] = sum(X[n, i, d1, ..., dk] ^ 2),
|
7
7
|
where max(0, c - floor((size - 1) / 2)) <= i <= min(C - 1, c + ceil((size - 1) / 2)).
|
8
8
|
Y[n, c, d1, ..., dk] = X[n, c, d1, ..., dk] / (bias + alpha / size * square_sum[n, c, d1, ..., dk] ) ^ beta
|
9
9
|
**Attributes**
|
10
10
|
* **alpha**:
|
11
11
|
Scaling parameter.
|
12
12
|
* **beta**:
|
13
13
|
The exponent.
|
14
14
|
* **bias**:
|
15
15
|
* **size** (required):
|
16
16
|
The number of channels to sum over
|
17
17
|
**Inputs**
|
18
18
|
* **X** (heterogeneous) - **T**:
|
19
19
|
Input data tensor from the previous operator; dimensions for image
|
20
20
|
case are (N x C x H x W), where N is the batch size, C is the number
|
21
21
|
of channels, and H and W are the height and the width of the data.
|
22
22
|
For non image case, the dimensions are in the form of (N x C x D1 x
|
23
23
|
D2 ... Dn), where N is the batch size. Optionally, if dimension
|
24
24
|
denotation is in effect, the operation expects the input data tensor
|
25
25
|
to arrive with the dimension denotation of [DATA_BATCH,
|
26
26
|
DATA_CHANNEL, DATA_FEATURE, DATA_FEATURE ...].
|
27
27
|
**Outputs**
|
28
28
|
* **Y** (heterogeneous) - **T**:
|
29
29
|
Output tensor, which has the shape and type as input tensor
|
30
30
|
**Type Constraints**
|
31
31
|
* **T** in (
|
32
|
-
tensor(bfloat16),
|
33
32
|
tensor(double),
|
34
33
|
tensor(float),
|
35
34
|
tensor(float16)
|
36
35
|
):
|
37
36
|
Constrain input and output types to float tensors.
|