Merge pull request #85 from HUJH511/main

Submission - LearnAFE: A Learnable Analog Front-End for Keyword Spotting

VLSI24/submitted_notebooks/LearnAFE/.gitignore

@@ -0,0 +1,18 @@
+# ---------- User-specified files to ignore ---------- 
+# Edit the following if necessary
+.conda/*
+AFE_Config/SpiceAC/*
+!AFE_Config/SpiceAC/README.md
+AFE_Config/SpiceTrans/*
+!AFE_Config/SpiceTrans/README.md
+SpeechCommands/*
+!SpeechCommands/README.md
+ckpts/*
+!ckpts/README.md
+logs/*
+!logs/README.md
+*.pyc
+*.pt
+*.pth
+*.tar.gz
+# ---------- End of user-specified list of files to ignore ---------- 

VLSI24/submitted_notebooks/LearnAFE/AFE_Config/Design/DSF16.txt

@@ -0,0 +1,7 @@
+C1_W	22
+C2_W	60
+I1	300
+I2	100
+scale	1.298
+C_spk	5
+Vth	1

VLSI24/submitted_notebooks/LearnAFE/AFE_Config/Design/README.md

@@ -0,0 +1 @@
+### Directory to place initial design file

VLSI24/submitted_notebooks/LearnAFE/AFE_Config/README.md

@@ -0,0 +1,10 @@
+# Directory to place configurations of analog front-end
+
+```
+./LearnAFE_KWS/AFE_Config
+|
+├───Design
+│   └───DSF16.txt
+├───SpiceAC
+└───SpiceTrans
+```

VLSI24/submitted_notebooks/LearnAFE/AFE_Config/SpiceAC/README.md

@@ -0,0 +1 @@
+### Directory to place spice files

VLSI24/submitted_notebooks/LearnAFE/AFE_Config/SpiceTrans/README.md

@@ -0,0 +1 @@
+### Directory to place spice files

VLSI24/submitted_notebooks/LearnAFE/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 HU JINHAI
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

VLSI24/submitted_notebooks/LearnAFE/README.md

@@ -0,0 +1,128 @@
+#   LearnAFE: A Learnable Analog Front-End for Keyword Spotting
+This repository contains code for keyword spotting with learnable analog front-end.
+
+
+##  Table of Content
+1. [Structure](#structure)
+2. [Getting Started](#getting-started)
+3. [Customization](#customization)
+4. [Architecture](#architecture)
+5. [Results](#results)
+6. [License](#license)
+7. [Acknowledgement](#acknowledgement)
+
+
+##  Structure
+While it is possible to have a stand-alone notebook, we have decided to split up the code into modules to improve readability and documentation. The following outline the directory structure.
+```
+./LearnAFE_KWS/
+|
+├───AFE_Config
+│   ├───Design
+│   ├───SpiceAC
+│   └───SpiceTrans
+├───ckpts
+├───Figures
+├───logs
+├───SpeechCommands
+│   └───speech_commands_v0.02
+├───src
+│   ├───data
+│   ├───models
+│   ├───utils
+│   ├───train.py
+│   └───test.py
+├───.gitignore
+├───environment.yaml
+├───README.md
+└───LearnAFE_KWS.ipynb
+```
+
+
+##  Getting Started
+
+### Environment Setup
+```
+./home/
+|
+├───EDA_Tools
+│   ├───magic
+│   └───open_pdks
+└───LearnAFE_KWS
+```
+Suggest to setup following environment under `EDA_Tools`
+```
+%%install ngspice and other possible configuration files
+cd /home/EDA_Tools
+sudo pip3 install flake8
+sudo pip3 install setuptools-scm
+sudo apt update
+sudo apt install build-essential tcl8.6 tcl8.6-dev tk8.6 tk8.6-dev libx11-dev libcairo2-dev libxpm-dev
+sudo apt update
+sudo apt install flex
+sudo apt install bison
+sudo apt install libcairo2-dev libjpeg-dev
+sudo apt install ngspice
+sudo apt install gwave
+sudo apt-get install xterm
+sudo apt-get install vim-gtk
+sudo apt-get install build-essential tcl-dev tk-dev libcairo2-dev libncurses-dev libx11-dev
+git clone https://github.com/RTimothyEdwards/magic.git
+cd magic
+./configure
+make
+sudo make install
+```
+Open PDK Installation
+```
+%%download and install open pdk
+cd /home/EDA_Tools
+git clone https://github.com/RTimothyEdwards/open_pdks
+cd open_pdks
+./configure --enable-sky130-pdk --enable-sram-sky130
+make
+sudo make install
+make veryclean
+```
+
+%% Python Setup
+The `environment.yml` can be used as a reference for the packages required. Before running the notebook, ensure that the circuit-design tools and the required packages have been installed successfully. Additionally, create a textfile in SpeechCommands -> speech_commands_v0.02 -> noise_list.txt with the following content. This is used to add background noise to the data for a more robust result.
+```
+%% noise_list.txt
+_background_noise_/doing_the_dishes.wav
+_background_noise_/dude_miaowing.wav
+_background_noise_/exercise_bike.wav
+_background_noise_/pink_noise.wav
+_background_noise_/running_tap.wav
+_background_noise_/white_noise.wav
+```
+
+
+##  Customization
+- Code-a-Chip
+    | Data Augmentation    | Filter Version | SR  | Hop length | Overlap | Max samples | Model          | Freeze |
+    |----------------------|----------------|-----|------------|---------|-------------|----------------|--------|
+    | Resample + Noise     | DSF16          | 20k | 640        | False   | 30          | LearnAFE_v2    | False  |
+
+
+##  Architecture
+- Learnable filter-bank on time domain
+
+    Waveform --> `START TRAIN` --> { Time domain filter } --> Filtered waveform --> { Spiking + Framing } --> Feature Map --> { DSCNN } --> Classification Result
+
+
+##  Results
+| Data Augmentation | Model         | AFE Config     | Freeze | SR   | Hop_len | Overlap | Train Acc | Valid Acc | Test Acc | 20dB | 5dB  |
+|-------------------|---------------|----------------|--------|------|---------|---------|-----------|-----------|----------|------|------|
+| Resample + Noise  | LearnAFE_v2   | Design/DSF16   | True   | 20k  | 640     | False   | 94.0      | 91.3      | 90.2     | 89.1 | 82.7 |
+| Resample + Noise  | LearnAFE_v2   | Design/DSF16   | False  | 20k  | 640     | False   | 95.6      | 93.8      | 92.7     | 92.4 | 89.4 |
+
+
+##  License
+This project is licensed under the MIT License. See LICENSE for more details.<br />
+Additionally the first version of this work has been accepted for publication in DAC. The following citation can be modified once the proceeding is made available.<br />
+J. Hu, Z. Zhang, C. S. Leow, W. L. Goh, and Y. Gao, “Late Breaking Results: Circuit-Algorithm Co-design for Learnable Audio Analog Front-End,” in 61st ACM/IEEE Design Automation Conf. (DAC), accepted, 2024.
+
+
+##  Acknowledgement
+This work was supported by the Agency for Science, Technology and Research (A*STAR), Singapore under the Nanosystems at the Edge programme, grant No. A18A1b0055. We thank Professor Zhengya Zhang for his insightful comments.

VLSI24/submitted_notebooks/LearnAFE/SpeechCommands/README.md

@@ -0,0 +1,12 @@
+# Directory to place Speech Commands Dataset
+
+A noise list file is added under `SpeechCommands/speech_commands_v0.02`
+```
+%% noise_list.txt
+_background_noise_/doing_the_dishes.wav
+_background_noise_/dude_miaowing.wav
+_background_noise_/exercise_bike.wav
+_background_noise_/pink_noise.wav
+_background_noise_/running_tap.wav
+_background_noise_/white_noise.wav
+```

VLSI24/submitted_notebooks/LearnAFE/ckpts/README.md

@@ -0,0 +1 @@
+# Directory to place temporary models/checkpoints

VLSI24/submitted_notebooks/LearnAFE/environment.yml

@@ -0,0 +1,181 @@
+name: /home/LearnAFE_KWS/.conda
+channels:
+  - conda-forge
+  - defaults
+dependencies:
+  - _libgcc_mutex=0.1
+  - _openmp_mutex=4.5
+  - asttokens=2.4.1
+  - blas=1.0
+  - brotli=1.0.9
+  - brotli-bin=1.0.9
+  - bzip2=1.0.8
+  - ca-certificates=2024.3.11
+  - comm=0.2.2
+  - contourpy=1.2.0
+  - cycler=0.11.0
+  - cyrus-sasl=2.1.28
+  - dbus=1.13.18
+  - debugpy=1.6.7
+  - decorator=5.1.1
+  - exceptiongroup=1.2.0
+  - executing=2.0.1
+  - expat=2.5.0
+  - fontconfig=2.14.1
+  - fonttools=4.25.0
+  - freetype=2.12.1
+  - glib=2.78.4
+  - glib-tools=2.78.4
+  - gst-plugins-base=1.14.1
+  - gstreamer=1.14.1
+  - icu=73.1
+  - importlib-metadata=7.1.0
+  - importlib_metadata=7.1.0
+  - ipykernel=6.29.3
+  - ipython=8.22.2
+  - jedi=0.19.1
+  - jpeg=9e
+  - jupyter_client=8.6.1
+  - jupyter_core=5.7.2
+  - kiwisolver=1.4.4
+  - krb5=1.20.1
+  - lcms2=2.12
+  - ld_impl_linux-64=2.38
+  - lerc=3.0
+  - libbrotlicommon=1.0.9
+  - libbrotlidec=1.0.9
+  - libbrotlienc=1.0.9
+  - libclang=14.0.6
+  - libclang13=14.0.6
+  - libcups=2.4.2
+  - libdeflate=1.17
+  - libedit=3.1.20230828
+  - libffi=3.4.4
+  - libgcc-ng=13.2.0
+  - libgfortran-ng=11.2.0
+  - libgfortran5=11.2.0
+  - libglib=2.78.4
+  - libgomp=13.2.0
+  - libiconv=1.16
+  - libllvm14=14.0.6
+  - libopenblas=0.3.21
+  - libpng=1.6.39
+  - libpq=12.17
+  - libsodium=1.0.18
+  - libstdcxx-ng=11.2.0
+  - libtiff=4.5.1
+  - libuuid=1.41.5
+  - libwebp-base=1.3.2
+  - libxcb=1.15
+  - libxkbcommon=1.0.1
+  - libxml2=2.10.4
+  - lz4-c=1.9.4
+  - matplotlib=3.8.0
+  - matplotlib-base=3.8.0
+  - matplotlib-inline=0.1.6
+  - munkres=1.1.4
+  - mysql=5.7.24
+  - ncurses=6.4
+  - nest-asyncio=1.6.0
+  - numpy=1.26.4
+  - numpy-base=1.26.4
+  - openjpeg=2.4.0
+  - openssl=3.2.1
+  - packaging=24.0
+  - parso=0.8.3
+  - pcre2=10.42
+  - pexpect=4.9.0
+  - pickleshare=0.7.5
+  - pillow=10.2.0
+  - pip=23.3.1
+  - platformdirs=4.2.0
+  - ply=3.11
+  - prompt-toolkit=3.0.42
+  - psutil=5.9.8
+  - ptyprocess=0.7.0
+  - pure_eval=0.2.2
+  - pygments=2.17.2
+  - pyparsing=3.0.9
+  - pyqt=5.15.10
+  - pyqt5-sip=12.13.0
+  - python=3.11.8
+  - python-dateutil=2.9.0
+  - python_abi=3.11
+  - pyzmq=25.1.2
+  - qt-main=5.15.2
+  - readline=8.2
+  - setuptools=68.2.2
+  - sip=6.7.12
+  - six=1.16.0
+  - sqlite=3.41.2
+  - stack_data=0.6.2
+  - tk=8.6.12
+  - tornado=6.4
+  - traitlets=5.14.2
+  - typing_extensions=4.10.0
+  - wcwidth=0.2.13
+  - wheel=0.41.2
+  - xz=5.4.6
+  - zeromq=4.3.5
+  - zipp=3.17.0
+  - zlib=1.2.13
+  - zstd=1.5.5
+  - pip:
+      - ax-platform==0.3.7
+      - botorch==0.10.0
+      - cffi==1.16.0
+      - cmake==3.29.0.1
+      - filelock==3.13.3
+      - fsspec==2024.3.1
+      - gpytorch==1.11
+      - ipywidgets==8.1.2
+      - jaxtyping==0.2.28
+      - jinja2==3.1.3
+      - joblib==1.3.2
+      - jupyterlab-widgets==3.0.10
+      - linear-operator==0.5.1
+      - lit==18.1.2
+      - markupsafe==2.1.5
+      - mpmath==1.3.0
+      - multipledispatch==1.0.0
+      - mypy-extensions==1.0.0
+      - networkx==3.2.1
+      - nvidia-cublas-cu12==12.1.3.1
+      - nvidia-cuda-cupti-cu12==12.1.105
+      - nvidia-cuda-nvrtc-cu12==12.1.105
+      - nvidia-cuda-runtime-cu12==12.1.105
+      - nvidia-cudnn-cu12==8.9.2.26
+      - nvidia-cufft-cu12==11.0.2.54
+      - nvidia-curand-cu12==10.3.2.106
+      - nvidia-cusolver-cu12==11.4.5.107
+      - nvidia-cusparse-cu12==12.1.0.106
+      - nvidia-nccl-cu12==2.19.3
+      - nvidia-nvjitlink-cu12==12.4.99
+      - nvidia-nvtx-cu12==12.1.105
+      - opt-einsum==3.3.0
+      - pandas==2.2.1
+      - plotly==5.20.0
+      - pycparser==2.21
+      - pyre-extensions==0.0.30
+      - pyro-api==0.1.2
+      - pyro-ppl==1.9.0
+      - pysoundfile==0.9.0.post1
+      - pytz==2024.1
+      - scikit-learn==1.4.1.post1
+      - scipy==1.12.0
+      - soundfile==0.12.1
+      - sympy==1.12
+      - tabulate==0.9.0
+      - tenacity==8.2.3
+      - threadpoolctl==3.4.0
+      - torch==2.2.1
+      - torchaudio==2.2.1
+      - torchinfo==1.8.0
+      - torchvision==0.17.1
+      - tqdm==4.66.2
+      - triton==2.2.0
+      - typeguard==2.13.3
+      - typing-inspect==0.9.0
+      - tzdata==2024.1
+      - widgetsnbextension==4.0.10
+prefix: /home/LearnAFE_KWS/.conda