Proverbs of Computer Vision and Robotics

August 20, 2013

Pretending I have been at this long enough to actually have some wisdom, here are a few popular proverbs and maxims recast to the fields of Computer Vision and Robotics:

  1. It is a bad researcher who blames the image database.
  2. A camera is not an eye.
  3. When all you have is a Hough Transform, everything looks like a line.
  4. A low resolution image analysed in real-time is worth two HD images in a database.
  5. Every recognition algorithm misses something.
  6. Great saliency algorithms find the same fixations.
  7. Check for obstacles before the robot moves.
  8. Give a radiologist a medical scan, and she will find the lesions. Teach a computer to do it, and it will be faster and cost less.
  9. Some days you spot the stairs, other days your robot drives down them.
  10. There’s no such thing as a perfect heuristic.
  11. Two bad images don’t make a good one.
  12. Two cameras give stereo depth, three give a headache.
  13. You can’t fight the granting agency.

Here are their more well-known versions (for the most part I tried to maintain a rough parity of meaning):

  1. It is a bad carpenter who blames his tools.
  2. A house is not a home.
  3. When all you have is a hammer, everything looks like a nail.
  4. A bird in the hand is worth two in the bush.
  5. Every rule has an exception.
  6. Great minds think alike.
  7. Look before you leap.
  8. Give a man a fish and you feed him for a day. Teach a man to fish and you feed him for a lifetime.
  9. Some days you get the bear, other days the bear gets you.
  10. There’s no such thing as a free lunch.
  11. Two wrongs don’t make a right.
  12. Two’s company, three’s a crowd.
  13. You can’t fight city hall.