Academia.eduAcademia.edu

Outline

Title
Abstract
Background
Sensory Data Router
Evaluation and Discussion
Experimental Setup
Conclusions
References

Human Behavior Analysis by Means of Multimodal Context Mining

Profile image of Muhammad Asif RazzaqMuhammad Asif RazzaqProfile image of Muhammad Asif RazzaqMuhammad Asif Razzaq
https://doi.org/10.3390/S16081264
visibility

…

description

19 pages

link

1 file

Abstract

There is sufficient evidence proving the impact that negative lifestyle choices have on people's health and wellness. Changing unhealthy behaviours requires raising people's self-awareness and also providing healthcare experts with a thorough and continuous description of the user's conduct. Several monitoring techniques have been proposed in the past to track users' behaviour; however, these approaches are either subjective and prone to misreporting, such as questionnaires, or only focus on a specific component of context, such as activity counters. This work presents an innovative multimodal context mining framework to inspect and infer human behaviour in a more holistic fashion. The proposed approach extends beyond the state-of-the-art, since it not only explores a sole type of context, but also combines diverse levels of context in an integral manner. Namely, low-level contexts, including activities, emotions and locations, are identified from heterogeneous sensory data through machine learning techniques. Low-level contexts are combined using ontological mechanisms to derive a more abstract representation of the user's context, here referred to as high-level context. An initial implementation of the proposed framework supporting real-time context identification is also presented. The developed system is evaluated for various realistic scenarios making use of a novel multimodal context open dataset and data on-the-go, demonstrating prominent context-aware capabilities at both low and high levels. The last World Health Organization (WHO) global status report on noncommunicable diseases reveals that illnesses associated with lifestyle choices are currently the leading cause of death worldwide [1]. As a matter of fact, non-communicable diseases are responsible for more than two-thirds of the world's deaths, with more than 40% of them representing premature deaths under the age of 70 years. Recognizing this seriously worrying epidemic scenario, the WHO has defined a clear roadmap to alter the course of the so-called " slow-moving public health disaster ". Most of the policies

Related papers

Health 2020 and Action Plan for Public Health – New Understanding of Public Health
paulina karwowska

Health Problems of Civilization, 2017

View PDFchevron_right
Priority actions for the non-communicable disease crisis
Thomas Gaziano

2011

The UN High-Level Meeting on Non-Communicable Diseases (NCDs) in September, 2011, is an unprecedented opportunity to create a sustained global movement against premature death and preventable morbidity and disability from NCDs, mainly heart disease, stroke, cancer, diabetes, and chronic respiratory disease. The increasing global crisis in NCDs is a barrier to development goals including poverty reduction, health equity, economic stability, and human security. The Lancet NCD Action Group and the NCD Alliance propose fi ve overarching priority actions for the response to the crisis-leadership, prevention, treatment, international cooperation, and monitoring and accountability-and the delivery of fi ve priority interventions-tobacco control, salt reduction, improved diets and physical activity, reduction in hazardous alcohol intake, and essential drugs and technologies. The priority interventions were chosen for their health eff ects, cost-eff ectiveness, low costs of implementation, and political and fi nancial feasibility. The most urgent and immediate priority is tobacco control. We propose as a goal for 2040, a world essentially free from tobacco where less than 5% of people use tobacco. Implementation of the priority interventions, at an estimated global commitment of about US$9 billion per year, will bring enormous benefi ts to social and economic development and to the health sector. If widely adopted, these interventions will achieve the global goal of reducing NCD death rates by 2% per year, averting tens of millions of premature deaths in this decade. NCD Alliance/International Diabetes Federation, Brussels, Belgium (R Colagiuri BEd,

View PDFchevron_right
The Second International Conference on Global Health Challenges GLOBAL HEALTH 2013 Technical Program Committee
Amy Fielden

in Lisbon, Portugal, continued a series of events taking a global perspective on population health, from national to crosscountry approaches, multiplatform technologies, from drug design to medicine accessibility; everything under mobile, ubiquitous, and personalized characteristics of new age population. Recent advances in technology and computational science influenced a large spectrum of branches in approaching population health. Despite significant progress, many challenges exist, including health informatics, crosscountry platforms interoperability, system and laws harmonization, protection of health data, practical solutions, accessibility to health services, and many others. Along with technological progress, personalized medicine, ambient assistance and pervasive health complement patient needs. A combination of classical and information-driven approach is developing now, where diagnosis systems, data protection mechanisms, remote assistance and hospital-processes are conver...

View PDFchevron_right
Empowering Citizens with Tools for Personalized Health is the Future of Effective Public Health Responses
Andrew Ahn

Advanced Sciences and Technologies for Security Applications, 2021

The global health security landscape is characterized as 'messes' and 'wicked problems' that proliferate in this age of complexity. In designing global health security solutions, '…simple, quick fix solutions that flounder in the face of interconnectedness, volatility and uncertainty' (Jackson in Critical Systems Thinking and the Management of Complexity. John Wiley and Sons, NJ, 2019 [1]) will not suffice and may create unintended consequences. For example, recent years have seen the threat presented by bio-risks heighten significantly. Infectious disease emergencies have included Severe Acute Respiratory Syndrome (SARS), Foot and Mouth Disease (FMD), Middle East Respiratory Syndrome (MERS), H1N1 influenza, Ebola, Zika and now COVID-19. These outbreaks have had significant global and public health impact. Considering the current COVID-19 pandemic, as of January 2021, the COVID-19 pandemic has seen upwards of 100 million cases, 2.1 million deaths, and has significantly impacted the global economy and affected the most vulnerable. The impacts have seen the closure of borders, economic disruptions and failures, strained and overwhelmed health care systems, failure of supply chains, all of which are contributing to a human and national security issue. Sensemaking is the activity that enables us to turn the ongoing complexity of the world into a 'situation that is comprehended explicitly in words and that serves as a springboard into action' (Weick et al. in Organ Sci 16:409-421, 2005 [2]). The importance of sensemaking as it pertains to individual and public health is that it enables us to plan, design and take action in real-time to events that affect or 'shock' our health and healthcare system. With advances in technology over the last decade, from robust cloud infrastructures, to artificial intelligence, to wearable devices, personalized health interventions are becoming increasingly accurate, accessible, and prevalent. As part of this shift, there has been a growing movement in the field of public health to include personalized medicine as an integrative element of public health which would have the potential to

View PDFchevron_right
VITAL TOPICS FORUM Chronic Disaster: Reimagining Noncommunicable Chronic Disease
Megan Carney, Abril Saldaña Tejeda

AMERICAN ANTHROPOLOGIST, Vol. 122, No. 3 , 2020

The connection between chronic diseases and largescale social forces is still absent from public policies designed to address obesity and diabetes (Tsai et al. 2017). Focusing on individual behavior or “ignorance” overlooks the histories of poverty and disadvantage while making women feel responsible not only for their own health but for the health of generations to come (Manderson 2016). Public health officials keep prescribing responsible citizens able to look after their own health and well-being. At the same time, everyday violence in Mexico is leaving many with no life to look after, with no body to nourish.

View PDFchevron_right
The Fourth International Conference on Global Health Challenges Committee GLOBAL HEALTH Steering Committee GLOBAL HEALTH 2015 Technical Program Committee
Nuno Garcia

a series of events taking a global perspective on population health, from national to crosscountry approaches, multiplatform technologies, from drug design to medicine accessibility, everything under mobile, ubiquitous, and personalized characteristics of new age population. Recent advances in technology and computational science influenced a large spectrum of branches in approaching population health. Despite significant progresses, many challenges exist, including health informatics, crosscountry platforms interoperability, system and laws harmonization, protection of health data, practical solutions, accessibility to health services, and many others. Along with technological progress, personalized medicine, ambient assistance and pervasive health complement patient needs. A combination of classical and information-driven approach is developing now, where diagnosis systems, data protection mechanisms, remote assistance and hospital-processes are converging. The conference had the ...

View PDFchevron_right
Community intervention for the emerging epidemic of non-communicable diseases
Hazel Bradley

South African Journal of Clinical …, 2007

View PDFchevron_right

References (53)

  1. WHO. Global Status Report on Noncommunicable Diseases 2014; Technical Report; World Health Organization: Geneva, Switzerland, 2014.
  2. Burke, L.E.; Swigart, V.; Turk, M.W.; Derro, N.; Ewing, L.J. Experiences of Self-Monitoring: Successes and Struggles During Treatment for Weight Loss. Qual. Health Res. 2009, 19, 815-828.
  3. Shephard, R.J. Limits to the measurement of habitual physical activity by questionnaires. Br. J. Sports Med. 2003, 37, 197-206.
  4. Swan, M. Health 2050: The realization of personalized medicine through crowdsourcing, the Quantified Self, and the participatory biocitizen. J. Pers. Med. 2012, 2, 93-118.
  5. Fitbit Surge. Available online: https://www.fitbit.com/surge (accessed on 30 November 2015).
  6. Jawbone Up. Available online: https://jawbone.com/up (accessed on 30 November 2015).
  7. Misfit Shine. Available online: http://www.misfitwearables.com/products/shine (accessed on 30 November 2015).
  8. Garmin Vivoactive. Available online: http://sites.garmin.com/en-US/vivo/vivofit/ (accessed on 1 April 2016).
  9. Empatica Embrace. Available online: https://www.empatica.com/product-embrace (accessed on 30 November 2015).
  10. Oresko, J.; Jin, Z.; Cheng, J.; Huang, S.; Sun, Y.; Duschl, H.; Cheng, A.C. A Wearable Smartphone-Based Platform for Real-Time Cardiovascular Disease Detection Via Electrocardiogram Processing. IEEE Trans. Inf. Technol. Biomed. 2010, 14, 734-740.
  11. Banos, O.; Villalonga, C.; Damas, M.; Gloesekoetter, P.; Pomares, H.; Rojas, I. PhysioDroid: Combining Wearable Health Sensors and Mobile Devices for a Ubiquitous, Continuous, and Personal Monitoring. Sci. World J. 2014, 2014, 1-11.
  12. Patel, S.; Mancinelli, C.; Healey, J.; Moy, M.; Bonato, P. Using Wearable Sensors to Monitor Physical Activities of Patients with COPD: A Comparison of Classifier Performance. In Proceedings of the 6th International Workshop on Wearable and Implantable Body Sensor Networks, Berkeley, CA, USA, 3-5 June 2009; pp. 234-239.
  13. Pierce, W.D.; Cheney, C.D. Behavior Analysis and Learning; Psychology Press: New York, NY, USA; London, UK, 2013.
  14. Pung, H.; Gu, T.; Xue, W.; Palmes, P.; Zhu, J.; Ng, W.L.; Tang, C.W.; Chung, N.H. Context-aware middleware for pervasive elderly homecare. IEEE J. Sel. Areas Commun. 2009, 27, 510-524.
  15. Xu, J.; Wang, Y.; Barrett, M.; Dobkin, B.; Pottie, G.; Kaiser, W. Personalized, Multi-Layer Daily Life Profiling through Context Enabled Activity Classification and Motion Reconstruction: An Integrated Systems Approach. IEEE J. Biomed. Health Inform. 2015, 20, 177-188.
  16. Gaggioli, A.; Pioggia, G.; Tartarisco, G.; Baldus, G.; Corda, D.; Cipresso, P.; Riva, G. A Mobile Data Collection Platform for Mental Health Research. Pers. Ubiquitous Comput. 2013, 17, 241-251.
  17. Apple HealthKit. Available online: https://developer.apple.com/healthkit/ (accessed on 30 November 2015).
  18. Google Fit. Available online: https://developers.google.com/fit/overview (accessed on 30 November 2015).
  19. SAMI, S. Available online: https://developer.samsungsami.io/sami/sami-documentation/ (accessed on 30 November 2015).
  20. Banos, O.; Amin, M.; Khan, W.; Ali, T.; Afzal, M.; Kang, B.; Lee, S. Mining Minds: An innovative framework for personalized health and wellness support. In Proceedings of the 2015 9th International Conference on Pervasive Computing Technologies for Healthcare (PervasiveHealth), Istanbul, Turkey, 20-23 May 2015; Volume 15.
  21. Villalonga, C.; Banos, O.; Ali Khan, W.; Ali, T.; Razzaq, M.A.; Lee, S.; Pomares, H.; Rojas, I. High-Level Context Inference for Human Behavior Identification. In International Work-Conference on Ambient Assisted Living an Active Ageing (IWAAL 2015), Lecture Notes in Computer Science; Cleland, I., Guerrero, L., Bravo, J., Eds.; Springer International Publishing: Puerto Varas, Chile, 2015; Volume 9455, pp. 164-175.
  22. Banos, O.; Amin, M.; Khan, W.; Hussain, M.; Afzal, M.; Kang, B.; Lee, S. The Mining Minds Digital Health and Wellness Framework. BioMed. Eng. OnLine 2016, 15, 165-186.
  23. Roggen, D.; Troester, G.; Lukowicz, P.; Ferscha, L.; del R. Millan, J.; Chavarriaga, R. Opportunistic Human Activity and Context Recognition. Computer 2013, 46, 36-45.
  24. Amin, M.B.; Banos, O.; Khan, W.A.; Muhammad Bilal, H.S.; Gong, J.; Bui, D.M.; Cho, S.H.; Hussain, S.; Ali, T.; Akhtar, U.; et al. On Curating Multimodal Sensory Data for Health and Wellness Platforms. Sensors 2016, 16, doi:10.3390/s16070980.
  25. Banos, O.; Galvez, J.M.; Damas, M.; Pomares, H.; Rojas, I. Window Size Impact in Human Activity Recognition. Sensors 2014, 14, 6474-6499.
  26. Preece, S.J.; Goulermas, J.Y.; Kenney, L.P.J.; Howard, D. A Comparison of Feature Extraction Methods for the Classification of Dynamic Activities From Accelerometer Data. IEEE Trans. Biomed. Eng. 2009, 56, 871-879.
  27. Cover, T.M.; Hart, P.E. Nearest neighbor pattern classification. IEEE Trans. Inf. Theory 1967, 13, 21-27.
  28. Banos, O.; Damas, M.; Pomares, H.; Prieto, A.; Rojas, I. Daily living activity recognition based on statistical feature quality group selection. Expert Syst. Appl. 2012, 39, 8013-8021.
  29. Banos, O.; Damas, M.; Pomares, H.; Rojas, I. On the Use of Sensor Fusion to Reduce the Impact of Rotational and Additive Noise in Human Activity Recognition. Sensors 2012, 12, 8039-8054.
  30. Banos, O.; Calatroni, A.; Damas, M.; Pomares, H.; Rojas, I.; Sagha, H.; del R Millan, J.; Troster, G.; Chavarriaga, R.; Roggen, D. Kinect=IMU? Learning MIMO Signal Mappings to Automatically Translate Activity Recognition Systems across Sensor Modalities. In Proceeding of the 2012 16th International Symposium on Wearable Computers (ISWC), Newcastle, Australia, 18-22 June 2012; pp. 92-99.
  31. Han, M.; Vinh, L.T.; Lee, Y.K.; Lee, S. Comprehensive Context Recognizer Based on Multimodal Sensors in a Smartphone. Sensors 2012, 12, 12588-12605.
  32. Chaaraoui, A.A.; Padilla-López, J.R.; Climent-Pérez, P.; Flórez-Revuelta, F. Evolutionary joint selection to improve human action recognition with RGB-D devices. Expert Syst. Appl. 2014, 41, 786-794.
  33. Han, J.; Shao, L.; Xu, D.; Shotton, J. Enhanced computer vision with microsoft kinect sensor: A review. IEEE Trans. Cybern. 2013, 43, 1318-1334.
  34. Soomro, K.; Zamir, A.R. Action recognition in realistic sports videos. In Computer Vision in Sports; Springer: New York, NY, USA, 2014; pp. 181-208.
  35. Zhang, H.; Parker Lynne, E. CoDe4D: Color-depth local spatio-temporal features for human activity recognition from RGB-D videos. IEEE Trans. Circ. Syst. Video Technol. 2016, 26, 541-555.
  36. Gupta, S.; Mooney, R.J. Using Closed Captions as Supervision for Video Activity Recognition. Available online: http://www.cs.utexas.edu/ ml/papers/gupta.aaai10.pdf (accessed on 9 August 2016).
  37. Bang, J.; Lee, S. Call Speech Emotion Recognition for Emotion based Services. J. KIISE Softw. Appl. 2014, 41, 208-213.
  38. Le, B.V.; Bang, J.; Lee, S. Hierarchical Emotion Classification using Genetic Algorithms. In Proceedings of the Fourth Symposium on Information and Communication Technology, New York, NY, USA, 5 December 2013.
  39. Cristianini, N.; Shawe-Taylor, J. An Introduction to Support Vector Machines and Other Kernel-Based Learning Methods; Cambridge University Press: New York, NY, USA, 2000.
  40. Naver Maps API. Available online: http://developer.naver.com/wiki/pages/MapAPI (accessed on 3 October 2016).
  41. Banos, O.; Damas, M.; Pomares, H.; Rojas, F.; Delgado-Marquez, B.; Valenzuela, O. Human activity recognition based on a sensor weighting hierarchical classifier. Soft Comput. 2013, 17, 333-343.
  42. Banos, O.; Galvez, J.M.; Damas, M.; Guillen, A.; Herrera, L.J.; Pomares, H.; Rojas, I.; Villalonga, C.; Hong, C.; Lee, S. Multiwindow Fusion for Wearable Activity Recognition. In International Work-Conference on Artificial Neural Networks; Rojas, I., Joya, G., Catala, A., Eds.; Springer International Publishing: Palma de Mallorca, Spain, 2015; Volume 9095, pp. 290-297.
  43. W3C OWL Working Group. OWL 2 Web Ontology Language: Document Overview (Second Edition);
  44. Carnielli, W.A. Systematization of finite many-valued logics through the method of tableaux. J. Symb. Log. 1987, 52, 473-493.
  45. Hall, M.; Frank, E.; Holmes, G.; Pfahringer, B.; Reutemann, P.; Witten, I.H. The WEKA data mining software: an update. ACM SIGKDD Explor. Newsl. 2009, 11, 10-18.
  46. Apache Jena. Available online: https://jena.apache.org/ (accessed on 3 October 2016).
  47. Brickley, D.; Guha, R.V. RDF Schema 1.1; W3C Recommendation. Available online: https://www.w3.org/ TR/rdf-schema/ (accessed on 25 February 2014).
  48. Harris, S.; Seaborne, A. SPARQL 1.1 (SPARQL Query Language for RDF);
  49. Sirin, E.; Parsia, B.; Grau, B.C.; Kalyanpur, A.; Katz, Y. Pellet: A practical OWL-DL reasoner. J. Web Semant. 2007, 5, 51-53.
  50. Richardson, L.; Ruby, S. RESTful Web Services; O'Reilly Media, Inc.: Sebastopol, CA, USA, 2008.
  51. Samsung Galaxy S5. Available online: http://www.samsung.com/us/search/searchMain?Dy=1&Nty=1& Ntt=samsung+galaxy+s5&3Acq_csrf_token=undefined (accessed on 3 October 2016).
  52. LG G Watch R. Available online: http://www.lg.com/us/smart-watches/lg-W110-g-watch-r (accessed on 8 November 2015).
  53. Bulling, A.; Blanke, U.; Schiele, B. A Tutorial on Human Activity Recognition Using Body-worn Inertial Sensors. ACM Comput. Surv. 2014, 46, 1-33.

Related papers

Mining human behavior for health promotion
Taqdir Ali

2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2015

The monitoring of human lifestyles has gained much attention in the recent years. This work presents a novel approach to combine multiple context-awareness technologies for the automatic analysis of people's conduct in a comprehensive and holistic manner. Activity recognition, emotion recognition, location detection, and social analysis techniques are integrated with ontological mechanisms as part of a framework to identify human behavior. Key architectural components, methods and evidences are described in this paper to illustrate the interest of the proposed approach.

View PDFchevron_right
Prevention and management of non-communicable disease: the IOC consensus statement, Lausanne 2013
Rodrigo Martinez

British Journal of Sports Medicine, 2013

Morbidity and mortality from preventable, noncommunicable chronic disease (NCD) threatens the health of our populations and our economies. The accumulation of vast amounts of scientific knowledge has done little to change this. New and innovative thinking is essential to foster new creative approaches that leverage and integrate evidence through the support of big data, technology, and design thinking. The purpose of this paper is to summarize the results of a consensus meeting on NCD prevention sponsored by the International Olympic Committee (IOC) in April, 2013. Within the context of advocacy for multifaceted systems change, the IOC's focus is to create solutions that gain traction within health care systems. The group of participants attending the meeting achieved consensus on a strategy for the prevention and management of chronic disease that includes the following:

View PDFchevron_right
Sensing the'Health State'of our Society
Katayoun Farrahi

2011

Mobile phones are a pervasive platform for opportunistic sensing of behaviors and opinions. We show that location and communication sensors can be used to model individual symptoms, long-term health outcomes, and diff usion of opinions in society. For individuals, phone-based features can be used to predict changes in health, such as common colds, influenza, and stress, and automatically identify symptomatic days. For longer-term health outcomes such as obesity, we fi nd that weight changes of participants are correlated with ...

View PDFchevron_right
Curbing the lifestyle disease pandemic: making progress on an interdisciplinary research agenda for law and policy interventions
Brigit Toebes

By 2030, noncommunicable diseases (NCDs) will be the leading cause of death in every region in the world. While law and policy have an important role to play in curbing this pandemic, our current understanding of how they can most effectively be used is still limited. This contribution identifies a number of gaps in current research and insists on an interdisciplinary research agenda between law, health science and international relations aimed at designing concrete proposals for laws and policies to curb the NCD pandemic, both globally and domestically.

View PDFchevron_right
Top ten public health challenges to track in 2022
Kenesh Dzhusupov

We identify ten public health challenges that need to be closely tracked in 2022. These challenges are COVID-19, inadequate human resources for health, poor health systems financing, conflict and humanitarian crises, mental health, poverty, climate change, the health of children, reproductive health issues, and the infodemic. These global priorities, based on opinion of experts and current evidence and literature, This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

View PDFchevron_right

Cited by

A Multimodal Deep Log-Based User Experience (UX) Platform for UX Evaluation
Jamil Hussain

Sensors (Basel, Switzerland), 2018

The user experience (UX) is an emerging field in user research and design, and the development of UX evaluation methods presents a challenge for both researchers and practitioners. Different UX evaluation methods have been developed to extract accurate UX data. Among UX evaluation methods, the mixed-method approach of triangulation has gained importance. It provides more accurate and precise information about the user while interacting with the product. However, this approach requires skilled UX researchers and developers to integrate multiple devices, synchronize them, analyze the data, and ultimately produce an informed decision. In this paper, a method and system for measuring the overall UX over time using a triangulation method are proposed. The proposed platform incorporates observational and physiological measurements in addition to traditional ones. The platform reduces the subjective bias and validates the user's perceptions, which are measured by different sensors thro...

View PDFchevron_right
mlCAF: Multi-Level Cross-Domain Semantic Context Fusioning for Behavior Identification
Muhammad Razzaq

Sensors (Basel, Switzerland), 2017

The emerging research on automatic identification of user's contexts from the cross-domain environment in ubiquitous and pervasive computing systems has proved to be successful. Monitoring the diversified user's contexts and behaviors can help in controlling lifestyle associated to chronic diseases using context-aware applications. However, availability of cross-domain heterogeneous contexts provides a challenging opportunity for their fusion to obtain abstract information for further analysis. This work demonstrates extension of our previous work from a single domain (i.e., physical activity) to multiple domains (physical activity, nutrition and clinical) for context-awareness. We propose multi-level Context-aware Framework (mlCAF), which fuses the multi-level cross-domain contexts in order to arbitrate richer behavioral contexts. This work explicitly focuses on key challenges linked to multi-level context modeling, reasoning and fusioning based on the mlCAF open-source ont...

View PDFchevron_right
Multi-sensor fusion based on multiple classifier systems for human activity identification
alo uzoma

Human-centric Computing and Information Sciences, 2019

Multimodal sensors in healthcare applications have been increasingly researched because it facilitates automatic and comprehensive monitoring of human behaviors, high-intensity sports management, energy expenditure estimation, and postural detection. Recent studies have shown the importance of multi-sensor fusion to achieve robustness, high-performance generalization, provide diversity and tackle challenging issue that maybe difficult with single sensor values. The aim of this study is to propose an innovative multi-sensor fusion framework to improve human activity detection performances and reduce misrecognition rate. The study proposes a multi-view ensemble algorithm to integrate predicted values of different motion sensors. To this end, computationally efficient classification algorithms such as decision tree, logistic regression and k-Nearest Neighbors were used to implement diverse, flexible and dynamic human activity detection systems. To provide compact feature vector represe...

View PDFchevron_right
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved