Any example of this being used in the wild? I feel like it could save with backend costs (especially for hobby projects) but the bandwidth/loading time might make it infeasible?
Also if you have any novel model it will be trivial to reverse engineer it. You gotta send the weights over to the client and they can just run tensorflow.js themselves right?
Also if you have any novel model it will be trivial to reverse engineer it. You gotta send the weights over to the client and they can just run tensorflow.js themselves right?