Table 1 Notations and their definitions

L

The number of convolutional layers

p

The overall pruning rate for all channels

C^l

The original total number of channels in

each layer, 1≤l≤L

\(\mathbf {w}^{l}_{k}\)

The convolutional kernel,

\(\mathbf {w}^{l}_{k}\in \mathbb {R}^{C^{l-1}\times 3\times 3},1\le l\le L, 1\le k \le C^{l}\)

H^l

The height of channels, 1≤l≤L

W^l

The width of channels, 1≤l≤L

\(\mathbf {z}^{l}_{k}\)

The feature map or channel,

\(\mathbf {z}^{l}_{k}\in \mathbb {R}^{H^{l}\times W^{l}},1\le l\le L, 1\le k \le C^{l}\)

f^l

The feature saliency, \(\mathbf {f}^{l}\in \mathbb {R}^{C^{l}}\)

\(f^{l}_{k}\in \mathbf {f}^{l}, 1\le k\le C^{l}\)

[a^l]_i

The remaining channels in each layer in

the i-th training epoch, 1≤l≤L

\(\Theta ^{l}_{k}\)

The evaluation on channels’ significance w.r.t.

a single mini-batch, 1≤l≤L,1≤k≤C^l

[ξ^l]_i

The layers’ significance evaluation in the i-th

training epoch, 1≤l≤L

J

The loss function adopted to evaluate the difference

between the observed values and the actual ones

ε

The smoothing factor

s

The proportion of redistributing channels