1
$\begingroup$

If I use a desktop PC with a GPU, how long it might take to train face recognition deep neural network on let's say dataset of 2.6 million images and 2600 identities? I guess it should depend on various properties (e.g., type of the DNN). But I am just looking for a rough estimation. Is it a matter of hours/days or years?

Thanks!

$\endgroup$
  • 2
    $\begingroup$ Does not have enough details to be commented upon.E.g. $O(n^2)$ complexity compared to $O(n)$ can make a problem go to a year from a day. $\endgroup$ – DuttaA May 25 at 3:51
  • $\begingroup$ Very much dependent on your processing power and the computational complexity per DuttaA's comment. $\endgroup$ – DukeZhou May 29 at 20:32
2
$\begingroup$

Training time depends on a lot of parameters. Some of them are:

  1. Size of each image (resolution)
  2. Color/Monochrome image (color image has 3 times data if you consider RGB image)
  3. Like you mentioned on the type of DNN. No. of layers of DNN. No. of neurons in each layer.
  4. Total no. of images in the dataset. (2.6 million here)
  5. GPU you are using (you didn't mention which GPU you are using. There are GPUs with wide range of capabilities, to predict the time, you need to know the exact specs of GPU).
  6. Training time also depends on RAM on your machine and also how fast host PC can transfer data to GPU for processing.
  7. Since you mentioned face recognition, i am assuming you are using CNN, but if you again use fully connected network, training time will change and will obviously increase manifold.

Your task of classification and your database is mostly similar to ILSVRC that uses imagenet database.

Making some reasonable assumptions for the parameters you didn't mention, i feel your task is similar to ILSVRC and i am predicting the training time will be a few days.

Below are the links which mention time of training for ILSVRC.

https://mxnet-tqchen.readthedocs.io/en/latest/tutorials/imagenet_full.html

Below are the details of imagenet database for your comparison

https://en.wikipedia.org/wiki/ImageNet

$\endgroup$
  • $\begingroup$ Thank you very much for this detailed answer! $\endgroup$ – John May 25 at 9:22
1
$\begingroup$

YELP Dataset (200k images) used to take 5 hr for training to identify Five (5) classes on GPU - Nvidia 1080 Ti with 11 GB RAM. So I guess in your case it will take days. Again it will depend on the type of your GPU configuration and type of Architecture you will be using.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.