• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ÇÐȸÁö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ÇÐȸÁö > µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)

µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) µµÇнÀÀ» »ç¿ëÇÑ µ¿ÀÛ ÀÎÁö ¸ðµ¨ÀÇ ¼º´É Çâ»ó
¿µ¹®Á¦¸ñ(English Title) Improving Human Activity Recognition Model with Limited Labeled Data using Multitask Semi-Supervised Learning
ÀúÀÚ(Author) À̼®·æ   Aria Ghora Prabono   Bernardo Nugroho Yahya   Seok-Lyong Lee  
¿ø¹®¼ö·Ïó(Citation) VOL 34 NO. 03 PP. 0137 ~ 0147 (2018. 12)
Çѱ۳»¿ë
(Korean Abstract)
±â°è ÇнÀÀ» ÅëÇÑ Àΰ£ µ¿ÀÛ ÀÎÁö (human activity recognition) ½Ã½ºÅÛ¿¡¼­ Áß¿äÇÑ ¿ä¼Ò´Â ÃæºÐÇÑ ¾çÀÇ ¶óº§ µ¥ÀÌÅÍ (labeled data)¸¦ È®º¸ÇÏ´Â °ÍÀÌ´Ù. ±×·¯³ª ¶óº§ µ¥ÀÌÅ͸¦ È®º¸ÇÏ´Â ÀÏÀº ¸¹Àº ºñ¿ë°ú ½Ã°£À» ÇÊ¿ä·Î ÇÑ´Ù. ¸Å¿ì ÀûÀº ¼öÀÇ ¶óº§ µ¥ÀÌÅ͸¦ °¡Áö°í ÀÖ´Â »õ·Î¿î ȯ°æ (Ÿ°Ù µµ¸ÞÀÎ)¿¡¼­ µ¿ÀÛ ÀÎÁö ½Ã½ºÅÛÀ» ±¸ÃàÇÏ´Â °æ¿ì, ±âÁ¸ÀÇ È¯°æ (¼Ò½º µµ¸ÞÀÎ)ÀÇ µ¥ÀÌÅͳª ÀÌ È¯°æ¿¡¼­ ÇнÀµÈ ºÐ·ù±â(classifier)¸¦ »ç¿ëÇÏ´Â °ÍÀº µµ¸ÞÀÎÀÌ ¼­·Î ´Ù¸£±â ¶§¹®¿¡ ¹Ù¶÷Á÷ÇÏÁö ¾Ê´Ù. ±âÁ¸ÀÇ ±â°è ÇнÀ ¹æ¹ýµéÀÌ ÀÌ·¯ÇÑ ¹®Á¦¸¦ ÇØ°áÇÒ ¼ö ¾øÀ¸¹Ç·Î ÀüÀÌ ÇнÀ (transfer learning) ¹æ¹ýÀÌ Á¦½ÃµÇ¾úÀ¸¸ç, ÀÌ ¹æ¹ý¿¡¼­´Â ¼Ò½º µµ¸ÞÀο¡¼­ È®º¸ÇÑ Áö½ÄÀ» È°¿ëÇÏ¿© Ÿ°Ù µµ¸ÞÀο¡¼­ÀÇ ºÐ·ù±â ¼º´ÉÀ» ³ôÀ̵µ·Ï ÇÏ°í ÀÖ´Ù. º» ³í¹®¿¡¼­´Â ´ÙÁß Å½ºÅ© ½Å°æ¸Á (multitask neural network)À» »ç¿ëÇÏ¿© ¸Å¿ì Á¦ÇÑµÈ ¼öÀÇ µ¥ÀÌÅ͸¸À¸·Î Á¤È®µµ°¡ ³ôÀº µ¿ÀÛ ÀÎÁö ºÐ·ù±â¸¦ »ý¼ºÇÏ´Â ÀüÀÌ ÇнÀ¹æ¹ýÀ» Á¦¾ÈÇÑ´Ù. ÀÌ ¹æ¹ý¿¡¼­´Â ¼Ò½º ¹× Ÿ°Ù µµ¸ÞÀÎ ºÐ·ù±âÀÇ ¼Õ½Ç ÇÔ¼ö ÃÖ¼ÒÈ­°¡ º°°³ÀÇ Å½ºÅ©·Î °£ÁֵȴÙ. Áï, ÇϳªÀÇ ½Å°æ¸ÁÀ» »ç¿ëÇÏ¿© µÎ ŽºÅ©ÀÇ ¼Õ½Ç ÇÔ¼ö¸¦ µ¿½Ã¿¡ ÃÖ¼ÒÈ­ÇÏ´Â ¹æ½ÄÀ¸·Î Áö½Ä ÀüÀÌ(knowledge transfer)°¡ ÀϾ°Ô µÈ´Ù. ¶ÇÇÑ, Á¦¾ÈÇÑ ¹æ¹ý¿¡¼­´Â ¸ðµ¨ ÇнÀÀ» À§ÇÏ¿© ºñÁöµµ ¹æ½Ä(unsupervised manner)À¸·Î ¶óº§ÀÌ ºÎ¿©µÇÁö ¾ÊÀº µ¥ÀÌÅ͸¦ È°¿ëÇÑ´Ù. ½ÇÇè °á°ú, Á¦¾ÈÇÑ ¹æ¹ýÀº ±âÁ¸ÀÇ ¹æ¹ý¿¡ ºñÇÏ¿© ÀÏ°üÀûÀ¸·Î ¿ì¼öÇÑ ¼º´ÉÀ» º¸¿©ÁÖ°í ÀÖ´Ù.
¿µ¹®³»¿ë
(English Abstract)
A key to a well-performing human activity recognition (HAR) system through machine learning technique is the availability of a substantial amount of labeled data. Collecting sufficient labeled data is an expensive and time-consuming task. To build a HAR system in a new environment (i.e., the target domain) with very limited labeled data, it is unfavorable to naively exploit the data or trained classifier model from the existing environment (i.e., the source domain) as it is due to the domain difference. While traditional machine learning approaches are unable to address such distribution mismatch, transfer learning approach leverages the utilization of knowledge from existing well-established source domains that help to build an accurate classifier in the target domain. In this work, we propose a transfer learning approach to create an accurate HAR classifier with very limited data through the multitask neural network. The classifier loss function minimization for source and target domain are treated as two different tasks. The knowledge transfer is performed by simultaneously minimizing the loss function of both tasks using a single neural network model. Furthermore, we utilize the unlabeled data in an unsupervised manner to help the model training. The experiment result shows that the proposed work consistently outperforms existing approaches.
Å°¿öµå(Keyword) ÀüÀÌ ÇнÀ   ´ÙÁß Å½ºÅ© ÇнÀ   µ¿ÀÛ ÀÎÁö   transfer learning   multitask learning   activity recognition  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå