vincenthwt

5 ms·

1. 31.▲
   by vincenthwt 2y ago

   The claim that "a tiny NN can learn edges better" is misleading. Classical algorithms like Canny or Sobel are specifically designed for edge detection, making them faster, more reliable, and easier to use in controlled environment
2. 32.▲
   by vincenthwt 2y ago

   If you're looking for simpler, faster, and reliable edge detection algorithms, traditional methods like Canny, Sobel, Prewitt, and Laplacian are excellent choices. For real-time applications or resource-constrained systems, simpler met
3. 33.▲
   by vincenthwt 2y ago

   Yes, those methods are old, but they’re explainable and much easier to debug or improve compared to the black-box nature of neural networks. They’re still useful in many cases.
4. 34.▲
   by vincenthwt 2y ago

   Since Sudoku is a logic-based game, why not use Prolog to create a solver? Example of Prolog code can be found here, https://www.metalevel.at/sudoku/ .