5 ms·
Using Deep Learning models trained on static images is usually not a good idea. But one can simply extend the model to use 3D convolutions instead of 2D, use ti
by bitL 7y ago
Using Deep Learning models trained on static images is usually not a good idea. But one can simply extend the model to use 3D convolutions instead of 2D, use time as 3rd axis and then feed a sequence of images for training, getting much better results out of it without the wavy effects.
- londons_explore 7y agoAnd have the amount of training data and time scale from n^2 to n^3... That's why nobody has achieved good results with this yet...
- bitL 7y agoOh yes, but that might be comparable to attention-based models for NLP, complexity-wise. Using C3D is quite common since 2015 actually: http://vlg.cs.dartmouth.edu/c3d/ http://vlg.cs.dartmouth.edu/c3d/ https://arxiv.org/pdf/1412.0767.pdf https://arxiv.org/pdf/1412.0767.pdf Training a great model is expensive these days, FB routinely pays 6 figures per training run.