report: Write abstract

Author: jonnor

TODO.md

@@ -5,26 +5,12 @@ Feedback needed
 
 ## TODO
 
-### Draft 3
-
-
 ### Final 1
 
-Abstract
-
-- Write it!
-- Send to OK for feedback
-
-Report
-
-- Add short captions to figures, so List of Figures looks nice
-
 Reprod
 
 - Tag a branch for submitted thesis
-- Upload models to GH
 - Results are from git commit `b49efa5dde48f9fd72a32eff4c751d9d0c0de712`
-- Include perftools Python script in appendix?
 
 Materials
 
@@ -34,6 +20,10 @@ Background
 
 - Make missing images
 
+Report
+
+- Add short captions to figures, so List of Figures looks nice
+
 
 ### Draft 5
 

report/abstract.latex

@@ -5,12 +5,37 @@
 
 \begin{abstract}
 
-Purpose/Motivation
-Methods
-Results
-Conclusions
+Noise is a growing problem in urban areas,
+and according to the WHO is the second environmental cause of health problems in Europe.
+Noise monitoring using Wireless Sensor Networks is
+being applied in order to understand and help mitigate these noise problems.
+It is desireable that these sensor systems, in addition to logging the sound level,
+can indicate what the likely sound source is.
+However transmitting audio to a cloud system for classification is
+energy-intensive and may cause privacy issues.
+It is also critical for widespread adoption and dense sensor coverage that
+individual sensor nodes are low-cost.
+Therefore we propose to perform the noise classification on the sensor node,
+using a low-cost microcontroller.
+
+Several Convolutional Neural Networks were designed for the
+STM32L476 low-power microcontroller using the Keras deep-learning framework,
+and deployed using the vendor-provided X-CUBE-AI inference engine.
+The resource budget for the model was set at maximum 50\% utilization of CPU, RAM and FLASH.
+10 model variations were evaluated on the Environmental Sound Classification task
+using the standard Urbansound8k dataset.
+
+The best models used Depthwise-Separable convolutions with striding for downsampling,
+and were able to reach 70.9\% mean 10-fold accuracy while consuming only 20\% CPU.
+To our knowledge, this is the highest reported performance on Urbansound8k using a microcontroller.
+One of the models was also tested on device,
+demonstrating classification of environmental sounds in real-time.
+
+These results indicate that it is computationally feasible to classify environmental sound
+on low-power microcontrollers.
+Further development should make it possible to create wireless sensor-networks
+for noise monitoring with on-edge noise source classification.
 
-FIXME: write it
 \end{abstract}
 
 \thispagestyle{empty}

report/references.bib

@@ -294,7 +294,14 @@ @article{BarcelonaNoiseMonitoring2018
   year={2018}
 }
 
-
+@article{EnvironmentalSoundRecognition2014,
+title={Environmental sound recognition: a survey},
+volume={3},
+DOI={10.1017/ATSIP.2014.12},
+journal={APSIPA Transactions on Signal and Information Processing}, publisher={Cambridge University Press}, author={Chachada, Sachin and Kuo, C.-C. Jay},
+year={2014},
+pages={e14}
+}
 
 @article{SourceClassificationSensors,
 title = "Environmental noise monitoring using source classification in sensors",

report/report.md

@@ -824,7 +824,7 @@ Scaling the output to a range of 0-1 or -1,1 is also sometimes done.
 These changes have the effect of removing amplitude variations,
 forcing the model to focus on the patterns of the sound regardless of amplitude.
 
-`TODO: image of normalized mel-spectrogram. Or feature distribution of datasets w/without normalization?`
+<!-- `TODO: image of normalized mel-spectrogram. Or feature distribution of datasets w/without normalization? -->
 
 
 ### Analysis windows
@@ -965,7 +965,9 @@ and ran in 242 ms on a low-end microcontroller (ARM Cortex M0+ at 48 Mhz).
 \newpage
 ## Environmental Sound Classification
 
-`TODO: introduction paragraph`
+Environmental Sound Classification, or Environmental Sound Recognition
+is the task of classifying environmental sounds or noises.
+It has been researched actively within the machine learning community at least since 2006[@EnvironmentalSoundRecognition2014].
 
 ### Datasets
 \label{chapter:datasets}