I wanted to start a broader ideas worth trying thread around Computer Vision.
A useful angle is limitations, but it also connects with privacy considerations. Different tools and workflows can produce very different results even when the goal is similar.
It would be interesting to compare practical examples, experiments, setups, or lessons that show where Computer Vision is genuinely useful and where it still needs improvement.
A useful angle is limitations, but it also connects with privacy considerations. Different tools and workflows can produce very different results even when the goal is similar.
It would be interesting to compare practical examples, experiments, setups, or lessons that show where Computer Vision is genuinely useful and where it still needs improvement.