EfficientWord-Net

EfficientWord-Net copied to clipboard

complex hotwords support #Current Model Limitations Discussion

8

Hi, Thanks for your helpful research. I wonder if the current model can handle complex hot words like "Hey Siri" or just handle one word, like "Siri"? My second question...

amoazeni75

Problem with Dependencies #Docker Support

3

Hello I left a comment on Reddit saying I would give it a go, and you said if I had a problem to log it here, so here I am,...

Balro76

OSError: [Errno -9981] Input overflowed

1

I've installed the python library with `pip install EfficientWord-Net` onto a raspberry pi 2 with recent raspbian lite. However if i run the demo with `python -m eff_word_net.engine` i'll get...

mKenfenheuer

Upgrade / Update

2

No support for Python 12, is that planned? tflite is no longer maintained for windows / mac, you have to use the full tensorflow. ``` Note: We no longer release...

dlroy88osu

#45 Update dependencies for Python 3.10+, add TensorFlow for Windows/Mac

1

#45 Update dependencies for Python 3.10+, add TensorFlow for Windows/Mac - Updated dependencies to support Python 3.10 and later versions, addressing the requirements of #45. - Dropped support for Python...

kiri-i

update deps

1

update dependencies to latest

jrsarath

Concatenate melspec graph + basemodel graph

2

Hi, first of all, thank you for sharing this awesome work! I noticed that the first model version was divided into two separate graphs `logmelcalc.tflite` and `baseModel.tflite`. Wondering if there...

kfengtee

AssertionError: Minimum of 4 sample datapoints is required

1

I made my ref using generate_reference but for some reason the _ref file it made always had an empty list, so I opened the generate_reference.py file to find out that...

ZXTube

automictic wake word detect without even sound

6

it detecting wake word on silence , specially when i 1 times , its repeat it self twice, any solution?

Siyamfahad

Is it possible to add noise or reverb into sample label to increase the recall ?

4

Hi, The performance of the model are really good when the voice is clean, however if the background is not clean with some noisy or room reverb, the recall rate...

Leeviber