Fix AudioSolver and Reimplemented ImageSolver

[MacKey-255]

Added the human factor in the SolverAudio. Improved speech system in the SolverAudio. "Try again later" message detection in SolverAudio. WitAI service added. Secundary_speech_service was added to avoid errors of the first voice service in SolverAudio. Simple system of proxies added through the json file. Added selector of the method to solve the captcha. Better error handling (Avoid connection errors and other common ones). Shipping button implemented (Solve captcha and submit form). Improved information provided by the records. Implementing the Yolov3 neural network in the SolverImage system. Reimplemented the SolverImage system. Neural network without training (default on Yolo v3) The network must be trained, I don't have a computer to do that. Prediction only to ('bus', 'car', 'bicycle', 'fire hydrant', 'traffic light', 'vehicles', 'motorcycle') Auto Download weight yolo v3 not testing (My connection is slow)

[SizRex]

[MacKey-255]

Men, Remember that you are using a proxy so the response time is slower. Try to time out set to 300 seconds. SAlu2s

[SizRex]

I try using localhost

[MacKey-255]

Men, try using it without a proxy and you'll see how it works. I tested it on a dedicated server and on my PC and everything is on wheels, the only thing missing is to train the neural network. SAlu2s