ML-KWS-for-MCU
                                
                                
                                
                                    ML-KWS-for-MCU copied to clipboard
                            
                            
                            
                        Keyword spotting on Arm Cortex-M Microcontrollers
act_max
What is act_max parameter?
separable_convolution2d Is it equivalent to ds-cnn? why add convolution2d in the back of separable_convolution2d model?
Hi , After i train and freeze to pb file by following commands: python2 train.py --model_architecture ds_cnn --model_size_info 5 64 10 4 2 2 64 3 3 1 1 64...
What is the required tensorflow version for the repo, 1.4 or higher ? Anyone update it to 2.0?
Hi all: when i use the cmd: **python train.py --model_architecture ds_cnn --model_size_info 5 64 10 4 2 2 64 3 3 1 1 64 3 3 1 1 64 3...
i have trained my model on 2 keyword but sometime it is detecting false when the pitch is high.why this is happening ? 2500 dataset per keyword and used dnn...
I run file label.wav in validation_list.txt with Accuracy differ paper. I run with architecture CNN, accuracy on F1-score weighted is 85.4% and DS-CNN, accuracy on F1-score weighted is 88%. Please...
i want to train on 5 keywords and i wanted to know how much training data should i need. and what model should i use to ignore negative words and...
I tried to create a new header file for the DS-CNN by running: ``` python train.py --model_architecture ds_cnn ... python fold_batchnorm.py --model_architecture ds_cnn ... python quant_test.py --model_architecture ds_cnn ... ```...
static q7_t const conv1_wt[CONV1_OUT_CH*CONV1_KX*CONV1_KY]; static q7_t const conv1_bias[CONV1_OUT_CH]; static q7_t const conv2_ds_wt[CONV1_OUT_CH*CONV2_DS_KX*CONV2_DS_KY]; static q7_t const conv2_ds_bias[CONV1_OUT_CH]; static q7_t const conv2_pw_wt[CONV2_OUT_CH*CONV1_OUT_CH]; static q7_t const conv2_pw_bias[CONV2_OUT_CH]; static q7_t const conv3_ds_wt[CONV2_OUT_CH*CONV3_DS_KX*CONV3_DS_KY]; static q7_t...