How to produce optimized images for two output labels dataset?

[taylorty]

First thank you for making this amazing toolbox. My problem is that when I tried to generate optimized images, I got index out of range error. I think it is because I only have two output labels for my training dataset instead of 2000+ labels in the case of ImageNet

Traceback (most recent call last): File "optimize_image.py", line 225, in main() File "optimize_image.py", line 220, in main brave = args.brave, skipbig = args.skipbig) File "/Users/Documents/caffe/deep-visualization-toolbox/optimize/gradient_optimizer.py", line 177, in run_optimize xx, results = self._optimize(params, x0) File "/Users/Documents/caffe/deep-visualization-toolbox/optimize/gradient_optimizer.py", line 229, in _optimize push_label = self.labels[params.push_unit[0]] IndexError: list index out of range

[abaybektursun]

Yep, getting the same error. I have only 2 classes too

[arikpoz]

@taylorty, @abaybektursun I've made the tool substantially more generic: all the hardcoded parts where replaced with new settings parameters, most are deduced automatically by analyzing the network definition file I suggest you recheck using the updated version of the tool

Latest version is in https://github.com/arikpoz/deep-visualization-toolbox