Quiz Space

Deep Learning for Computer Vision · End Term · 13 Sept 2026 · May 2026 term

Deep Learning for Computer Vision End Term 13 Sept 2026 — Question 26

Question 26

+3 marksOne or more correct options

Select all correct statements.

Select all that apply.

  1. A

    Soft attention can form a weighted combination of visual features.

  2. B

    Self-attention allows token representations to depend directly on other tokens.

  3. C

    Vision Transformers adapt transformer-style token processing to visual inputs.

  4. D

    Transformer-based segmentation is restricted to producing one class label for an entire image.

Show answer

Correct answers

  • A

    Soft attention can form a weighted combination of visual features.

  • B

    Self-attention allows token representations to depend directly on other tokens.

  • C

    Vision Transformers adapt transformer-style token processing to visual inputs.

See the whole paperTake it as a mock test

Question 26 of 37 in the IIT Madras BS Deep Learning for Computer Vision (Deep Learning for Computer Vision) End Term paper sat on 13 Sept 2026, in the May 2026 term (Deep Learning For Computer Vision 13 Sep 26). It carries 3 marks.

More questions from this paper

  1. Q1An RGB image of size 128×96 is stored with 32-bit floating-point values for every channel. Ignoring metadata, how many …
  2. Q2Let a 1-D filter be h=[2,-1,3]. A library routine performs cross-correlation. Which filter should be supplied to the ro…
  3. Q3At a point on an ideal straight intensity edge, the local image-gradient vector is most naturally interpreted as pointi…
  4. Q4An image is smoothed successively with Gaussian kernels of standard deviations 2 and 3. Ignoring discretization, the eq…
  5. Q5Which design choice is most directly responsible for the rotation robustness of a SIFT descriptor?
  6. Q6A hidden activation h is used by two downstream branches, and both branches affect the same scalar loss L. During backp…
  7. Q7Suppose successive gradients keep pointing in a similar direction over several updates. Compared with plain gradient de…
  8. Q8Training loss continues to decrease, but validation loss has begun to increase consistently. Which interpretation is mo…
  9. Q9A 3×3 convolution maps 16 input channels to 32 output channels. If the input spatial size changes from 64×64 to 128×128…
  10. Q10A convolution kernel is reused at many spatial locations. The gradient of the loss with respect to one kernel coefficie…
  11. Q11Figure question
  12. Q12A depthwise-separable convolution first applies a spatial filter independently to each input channel and then uses a 1×…
  13. Q13When adapting a pretrained CNN to a related task with a relatively small labeled dataset, which training choice is gene…
  14. Q14Which processing pattern is characteristic of a two-stage object detector?
  15. Q15For semantic image segmentation with C classes, the model output before the final class decision is naturally organized…
  16. Q16Compared with a basic recurrent update, the gating mechanisms in LSTMs and GRUs are primarily intended to:
  17. Q17In self-attention, changing the query vector for one token while keeping keys and values fixed directly changes:
  18. Q18A standard way to convert an image for processing by a Vision Transformer is to:
  19. Q19Which statement correctly distinguishes the two model families?
  20. Q20Classifier-free guidance during sampling is based on combining:
  21. Q21Select all correct statements
  22. Q22Select all correct statements .
  23. Q23Select all correct statements.
  24. Q24Select all correct statements.
  25. Q25Select all correct statements .
  26. Q27Select all correct statements.
  27. Q28Select all correct statements .
  28. Q29An input feature map has spatial size 63×63. A 5×5 convolution is applied with stride 2 and padding 1. Compute the outp…
  29. Q30An image is smoothed successively by Gaussian kernels with σ1=1.2 and σ2=1.6. Ignoring discretization, compute the equi…
  30. Q31Figure question
  31. Q32A convolutional layer uses 3×3 kernels, 16 input channels and 32 output channels. The layer has one bias per output cha…
  32. Q33A depthwise-separable convolution receives 32 channels. It uses a 3×3 depthwise convolution followed by a 1×1 pointwise…
  33. Q34Figure question
  34. Q35Figure question
  35. Q36Figure question
  36. Q37Figure question