Quiz Space

Deep Learning for Computer Vision · End Term · 31 Aug 2025 · May 2025 term · Set QDB1

Question 16: Why does DETR typically exhibit poor performance in dete…

Question 16

+2 marksOne correct option

Why does DETR typically exhibit poor performance in detecting small objects compared to larger ones?

  1. A

    The CNN backbone used in DETR has a fixed receptive field that’s too large for small objects

  2. B

    The global self-attention mechanism in DETR tends to dilute the signal from small objects across all image locations

  3. C

    DETR’s loss function actively discards all detections smaller than 50x50 pixels

  4. D

    The object queries in DETR are programmed to ignore any object smaller than 10

  5. E

    DETR’s encoder-decoder architecture was specifically designed to only detect large objects

Show answer

Correct answer

  • B

    The global self-attention mechanism in DETR tends to dilute the signal from small objects across all image locations

See the whole paperTake it as a mock test

Question 16 of 52 in the IIT Madras BS Deep Learning for Computer Vision (Deep Learning for Computer Vision) End Term paper sat on 31 Aug 2025, in the May 2025 term (IIT M IMPROVEMENT AN EXAM QIA3 31 Aug 2025). It carries 2 marks.

More questions from this paper

  1. Q1Choose the correct matching:
  2. Q2Figure question
  3. Q3Identify the correct sequence of steps in a Canny edge detection pipeline. Steps are listed below: 1. Compute gradient …
  4. Q4Figure question
  5. Q5Match the derivative of activation functions f(x) with their counterparts on the right column accordingly. | 1) Leaky R…
  6. Q6Which of the following is the correct sequence of steps of the SIFT algorithm?\ Using the Taylor series expansion of th…
  7. Q7Figure question
  8. Q8Which one of the following statements is true?
  9. Q9Figure question
  10. Q10Which one of the following statements regarding hyperparameter tuning is false?
  11. Q11Figure question
  12. Q12Figure question
  13. Q13Figure question
  14. Q14What makes the Segment Anything Model (SAM) particularly advantageous for image annotation tasks compared to convention…
  15. Q15Statement 1: The Segment Anything Model (SAM) demonstrates zero-shot generalization capabilities across various downstr…
  16. Q17Which of the following statements are false? (Select all that apply)
  17. Q18Which of the following statements are true? (Select all possible correct options)
  18. Q19Which of the following statements about CLIP are TRUE? (Select ALL that apply)
  19. Q20Which one of the following statements is true:
  20. Q21Which of the following are examples of a high-pass filter?
  21. Q22Which of the following statements are true? (Select all that apply)
  22. Q23Which of the following techniques help control the exploding or vanishing gradient problem in recurrent neural networks?
  23. Q24Given is a 3 \times 3 8-bit grayscale image: \begin{bmatrix} 50 & 70 & 120 \ 90 & 30 & 80 \ 40 & 60 & 110 \end{bmatrix}…
  24. Q25Consider the grayscale image shown below as a 5 \times 5 matrix: \begin{bmatrix} 20 & 30 & 25 & 30 & 40 \ 45 & 10 & 40 …
  25. Q26Figure question
  26. Q27Figure question
  27. Q28Figure question
  28. Q29Figure question
  29. Q30Figure question
  30. Q31Figure question
  31. Q32Element 1: _
  32. Q33Element 2: _
  33. Q34Element 3: _
  34. Q35Element 4: _
  35. Q36A 4-dimensional input vector x = [5, 3, -1, 2] is passed to a hidden layer with a single neuron and an activation funct…
  36. Q37A 4-dimensional input vector x = [5, 3, -1, 2] is passed to a hidden layer with a single neuron and an activation funct…
  37. Q38A 4-dimensional input vector x = [5, 3, -1, 2] is passed to a hidden layer with a single neuron and an activation funct…
  38. Q39A 4-dimensional input vector x = [5, 3, -1, 2] is passed to a hidden layer with a single neuron and an activation funct…
  39. Q40A 4-dimensional input vector x = [5, 3, -1, 2] is passed to a hidden layer with a single neuron and an activation funct…
  40. Q41A 4-dimensional input vector x = [5, 3, -1, 2] is passed to a hidden layer with a single neuron and an activation funct…
  41. Q42Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…
  42. Q43Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…
  43. Q44Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…
  44. Q45Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…
  45. Q46Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…
  46. Q47Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…
  47. Q48Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…
  48. Q49Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…
  49. Q50Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…
  50. Q51Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…
  51. Q52Consider the BLIP model architecture with an input (image-text pair) as shown below: For each of the given entities, en…