Question text from the original paper, with its maths as pictures Figure from the passage in the original paper Question text from the original paper, with its maths as pictures

(A) Question text from the original paper, with its maths as pictures

(B) The pointwise layer uses a 1×1 kernel to mix information across channels after the depthwise step.

(D) The compression ratio achieved depends on both the kernel size and the number of output channels.

(C) Replacing a standard convolution with a depthwise-separable convolution always improves model accuracy.

Deep Learning Practice · End Term · 13 Sept 2026 · May 2026 term · Set 1

Deep Learning Practice End Term 13 Sept 2026 — Question 12

Question 12

+2 marksOne or more correct options

Select all that apply.

A
B
The pointwise layer uses a 1×1 kernel to mix information across channels after the depthwise step.
C
Replacing a standard convolution with a depthwise-separable convolution always improves model accuracy.
D
The compression ratio achieved depends on both the kernel size and the number of output channels.

Show answer

Correct answers

A
B
The pointwise layer uses a 1×1 kernel to mix information across channels after the depthwise step.
D
The compression ratio achieved depends on both the kernel size and the number of output channels.

See the whole paper Take it as a mock test

Question 12 of 31 in the IIT Madras BS Deep Learning Practice (Deep Learning Practice) End Term paper sat on 13 Sept 2026, in the May 2026 term (Deep Learning Practice 13 Sep 26 (Session 2)). It carries 2 marks.

Deep Learning Practice End Term 13 Sept 2026 — Question 12

Question 12

More questions from this paper