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

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Current Result Document : 232 / 232

ÇѱÛÁ¦¸ñ(Korean Title) ¿Â µð¹ÙÀ̽º °´Ã¼ °ËÃâÀ» À§ÇÑ »óȲ ÀÎÁö ±â¹ÝÀÇ ¸ðµ¨ ¼±Åà ±â¹ý
¿µ¹®Á¦¸ñ(English Title) Context-aware Model Selection Scheme for On-device Object Detection
ÀúÀÚ(Author) ÀÌÁظ𠠠¹Ú»óÇö   ¹®¼ö¹¬   Junmo Lee   Sanghyeon Park   Soo-Mook Moon   °­¼ºÁÖ   Á¤Ã¤Àº   Á¤±¤¼ö   Seongju Kang   Chaeeun Jeong   Kwangsue Chung  
¿ø¹®¼ö·Ïó(Citation) VOL 27 NO. 08 PP. 0388 ~ 0393 (2021. 08)
Çѱ۳»¿ë
(Korean Abstract)
DNN (Deep Neural Network)Àº ¾ó±¼ ÀνÄ, À½¼º ÀνÄ, °´Ã¼ °ËÃâ µî Áö´ÉÇü ¾îÇø®ÄÉÀ̼ǿ¡ ÇÊ¿äÇÑ ÄÄÇ»ÅÍ ºñÀü ±â¼úÀÌ´Ù. DNN¸ðµ¨Àº ¼ö¸¹Àº Àº´ÐÃþ°ú ÇнÀ ÆĶó¹ÌÅ͸¦ Æ÷ÇÔÇϱ⠶§¹®¿¡ °´Ã¼ °ËÃâ ¾Ë°í¸®ÁòÀº ¸¹Àº ¿¬»ê ¸®¼Ò½º¸¦ ¿ä±¸ÇÑ´Ù. ±×·¯¹Ç·Î, ¸ð¹ÙÀÏ ÀåÄ¡¿Í °°ÀÌ ¸®¼Ò½º°¡ Á¦ÇÑµÈ È¯°æ¿¡¼­´Â °´Ã¼ °ËÃâÀ» ¼öÇàÇϴµ¥ ¾î·Á¿òÀÌ ÀÖ´Ù. º» ³í¹®¿¡¼­´Â ¿Â µð¹ÙÀ̽º °´Ã¼ °ËÃâÀ» À§ÇÑ »óȲ ÀÎÁö ±â¹ÝÀÇ ¸ðµ¨ ¼±Åà ±â¹ýÀ» Á¦¾ÈÇÑ´Ù. ¸ðµ¨ÀÇ ¿¬»ê ¸®¼Ò½º¸¦ ÁÙÀ̱â À§ÇÏ¿© ½Ã°ø°£ µµ¸ÞÀÎ º° OOI (Object Of Interest) ±×·ìÀ» Á¤ÀÇÇÏ°í °¢ µµ¸ÞÀο¡ ´ëÇØ °æ·®È­ ¸ðµ¨À» ÇнÀ½ÃŲ´Ù. Á¦¾ÈÇÏ´Â ±â¹ýÀº »óȲ Á¤º¸¸¦ ±â¹ÝÀ¸·Î ½Ã°ø°£ µµ¸ÞÀÎÀ» °áÁ¤ÇÏ°í ªÀº Áö¿¬ ½Ã°£À¸·Î Á¤È®ÇÑ ¹°Ã¼ °¨Áö°¡ °¡´ÉÇϵµ·Ï ÃÖÀûÀÇ DNN ¸ðµ¨À» ¼±ÅÃÇÑ´Ù. ±âÁ¸ÀÇ °´Ã¼ °ËÃâ ±â¹ý°úÀÇ ºñ±³ ½ÇÇèÀ» ÅëÇØ Á¦¾ÈÇÏ´Â ±â¹ýÀÌ À¯»çÇÑ Á¤È®µµ¸¦ ´Þ¼ºÇϸ鼭 ªÀº Áö¿¬ ½Ã°£À¸·Î °´Ã¼ °ËÃâÀÌ °¡´ÉÇÑ °ÍÀ» È®ÀÎÇÏ¿´´Ù.
¿µ¹®³»¿ë
(English Abstract)
The deep neural network (DNN) is computer vision technology required for intelligent applications such as face recognition, voice recognition, and object detection, etc. Since the DNN model contains a number of hidden layers and learning parameters, the object detection algorithm requires many computational resources. Therefore, it is difficult to perform object detection in resourceconstrained environments such as mobile devices. In this paper, we proposed a context-aware model selection scheme for on-device object detection. To reduce the computational resources of the DNN model, we defined the object of interest (OOI) groups for each spatiotemporal domain and trained lightweight models for each domain. The proposed scheme determines the spatiotemporal domain based on the context information and selects the optimal DNN model to enable accurate object detection with a short latency.
Å°¿öµå(Keyword) ÅõÇ¥ ½Ã½ºÅÛ   Á¦°ö ÅõÇ¥   ½Ãºô °ø°Ý   ½Å·Ú ½ÇÇà ȯ°æ   voting system   quadratic voting   sybil attack   trusted execution environment   DNN (Deep Neural Network)   °´Ã¼ °ËÃâ   ¿Â µð¹ÙÀ̽º   OOI (Object Of Interest)   ½Ã°ø°£ µµ¸ÞÀΠ  DNN (Deep Neural Network)   object detection   On-device   OOI (Object Of Interest)   spatiotemporal domain  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå