Resiliency of Deep Neural Networks under Quantization

Understand

The complexity of deep neural network algorithms for hardware implementation can be much lowered by optimizing the word-length of weights and signals.

• Direct quantization of floating-point weights, however, does not show good performance when the number of bits assigned is small.
• Retraining of quantized networks has been developed to relieve this problem.
• In this work, the effects of retraining are analyzed for a feedforward deep neural network (FFDNN) and a convolutional neural network (CNN).