4 ms·
Interesting idea, but would this not be relatively easy to crack, given that the problem would basically resume to mapping images to words (i.e. no OCR involved
by mtarnovan 18y ago
Interesting idea, but would this not be relatively easy to crack, given that the problem would basically resume to mapping images to words (i.e. no OCR involved, just a histogram would suffice) ?
- forkqueue 18y agoMy thoughts exactly. With the current implementation, you wouldn't even need anything as complex as a histogram, an MD5 dictionary of image files would be fine.