search

akarshzingade / image-similarity-deep-ranking

369 stars 103 forks source link

Results #4

Closed jodevak closed 6 years ago

jodevak commented 6 years ago

Thank you for this work, especially you blog post was very clear. I have a question, have you tired runng your code, if yes, can you provide a test cases, success measures or anything similar?

akarshzingade commented 6 years ago

Hey! One way to measure success is to compare the features of 2 images by computing their euclidean distance. Ideally, if you have trained the model with gap parameter 'g' set to 1, the euclidean distance of <1 means similar image and >1 means dissimilar image. Hope this helps!