Classification
Accuracy
Prediction
Recall
F1
Crying/Non-Crying
97.33%
97.44%
97.37%
97.33%
Hungry/Pain/Sleepy
81.08%
82.27%
80.77%
80.95%
Hungry/Pain/Sleepy/
Non-Crying
90.67%
86.70%
85.58%
85.71%
We can see that when judging whether the signal depicts a
crying baby, the SVM achieves a great result. The classification
of different reasons of why the baby is crying are also high. In
the comprehensive analysis, 90.97% accuracy is achieved.
VI.
C
ONCLUSION
In this paper we introduced the system architecture,
workflow and system implementation of a smart crib control
system. The system explores the idea of approaching crying
analysis as a sentiment analysis task. We use framing, endpoint
detection and cry unit detection to extract data signals. Then, we
extract feature vectors and use SFFS for feature selection.
Figure 6. Results after Pre-Processing steps
Finally, we put the final feature vector into the SVM that
implements o-v-o strategy to classify and predict.
At present, we have implemented a laboratory demo of the
smart crib and proposed a design of the smart crib system. As
future work, we will perform additional experiments using
larger datasets, which would also allow the application of more
sophisticated methodsq.
A
CKNOWLEDGMENT
This research was supported by the National Natural Science
Foundation of China (Grant No. 61202085), and the
fundamental research funds for the central university (Grant No.
N161704003).
R
EFERENCES
[1]
T. G. Dietterich and E. J. Horvitz. “Rise of concerns about AI: Reflections
and directions
,”
Communications of the ACM
, vol. 58, pp. 38-40, 2015.
[2]
B. L. R. Stojkoska and K. V. Trivodaliev. ”A review of Internet of Things
for smart home: Challenges and solutions
”, Journal of Cleaner
Production
, 140(3):1454-1464, 2017.
[3]
A. D. Floarea and V. Sgarciu. "Smart refrigerator: A next generation
refrigerator connected to the IoT," In 2016 8th International Conference
on Electronics, Computers and Artificial Intelligence, Ploiesti, pp. 19-24,
2016.
[4]
“SNOO-Happiest baby,” Happiest Baby, 2018. [Online]. Available:
https://www.happiestbaby.com/pages/snoo. [Accessed 12 4 2018].
[5]
S. Sharma and M. Tomar. Principles Of Growth And Development, p.
128, Isha Books, 2005.
[6]
A. P. Shahnawaz, "Sentiment analysis: approaches and open issues". In
2017 International Conference on Computing, Communication and
Automation (ICCCA), pp.154-158, 2017.
[7]
C. Severance. "Eben Upton: Raspberry Pi," Computer, 46(10):14-16,
2013.
[8]
"MySQL," [Online]. Available: https://www.mysql.com/. [Accessed 12
04 2018].
[9]
C. Wilson, T. Hargreaves, and R. Hauxwell-Baldwin. “Smart Homes and
their users: A systematic analysis and key challenges
”. Personal and
Ubiquitous Computing
, 19(2):463-476, 2015.
[10]
Y. Abdulaziz, S. M. Syed Ahmad. “Infant cry recognition system: A
comparison of system performance based on Mel frequency and linear
prediction cepstral coefficients," In Proceedings of the 2010 international
conference on information retrieval and knowledge management, pp.
260-263, 2010.
[11]
A. Poornima. “Basic Characteristics of Speech Signal Analysis”
International Journal of Innovative Research & Development, 5(4):
169-
173, 2016.
[12]
P. Pudil, F. J. Ferri, J. Novovicova, and J. Kittler, J. “Floating search
methods for feature selection with nonmonotonic criterion functions
”.
Pattern Recognition, 2, 279
–283, 1994.
[13]
W. J. Chen, H. Guo, R.A. Renaut, and K. Chen. "A new SVM model for
classifying genetic data," In 2010 International Conference on
Bioinformatics,
Computational
Biology,
Genomics
and
Chemoinformatics (BCBGC-10), pp.54-60, 2010.
[14]
K. Gaunt, J. Nacsa, and M. Penz. “Baby lucent: Pitfalls of applying
quantified self to baby products
”. In CHI EA’14, pp. 263-268, 2014.
Dostları ilə paylaş: |