ÆíÁý : 2024.3.28 ¸ñ 17:13
´º½º
³×À̹ö, ¡®ÀÎÅͽºÇÇÄ¡ 2021¡¯¿¡ AI 9°³ ³í¹® ¹ßÇ¥¡®¶óÀΡ¯ ¹ßÇ¥ ÇÕÇϸé 14°³ ´ÞÇØ¡¦¸®´õ½Ê È®º¸ ¹ÚÂ÷
±èµ¿±â ±âÀÚ  |  kdk@bikorea.net
ÆùƮŰ¿ì±â ÆùÆ®ÁÙÀ̱â ÇÁ¸°Æ®Çϱ⠸ÞÀϺ¸³»±â ½Å°íÇϱâ
½ÂÀÎ 2021.09.06  09:33:55
Æ®À§ÅÍ ÆäÀ̽ººÏ ¹ÌÅõµ¥ÀÌ ¿äÁò ³×À̹ö ±¸±Û msn

³×À̹ö(´ëÇ¥ ÇѼº¼÷)´Â Áö³­ 8¿ù 30ÀϺÎÅÍ 9¿ù 3ÀϱîÁö ÁøÇàµÈ À½¼º ¹× ½Åȣó¸® ÇÐȸ ‘ÀÎÅͽºÇÇÄ¡(INTERSPEECH) 2021’¿¡¼­ 9°³ÀÇ ³í¹®À» ¹ßÇ¥Çß´Ù°í 6ÀÏ ÀüÇß´Ù. 

ƯÈ÷, ³×À̹ö ÀϺ» °ü°è»ç ‘¶óÀÎ’ÀÌ ¹ßÇ¥ÇÑ ³í¹®À» ÇÕÇϸé ÃÑ 14°³¿¡ ´ÞÇÑ´Ù.

ÀÌ¿Í °°Àº ¼º°ú¸¦ ±â¹ÝÀ¸·Î, ³×À̹ö´Â ±Û·Î¹ú AI ¸®´õ½Ê È®º¸¿¡ Àû±ØÀûÀ¸·Î ³ª¼³ °èȹÀÌ´Ù.

¿ÃÇØ 22ȸ¸¦ ¸ÂÀº ‘ÀÎÅͽºÇÇÄ¡’´Â ±¸±Û, ÆäÀ̽ººÏ, ¾Æ¸¶Á¸, ¾Ë¸®¹Ù¹Ù µî ±Û·Î¹ú ±â¾÷ÀÇ À½¼ºÀÎ½Ä °ü·Ã ÃֽŠ±â¼úÀÌ °øÀ¯µÇ´Â ÀÚ¸®·Î, ICASSP¿Í ÇÔ²² À½¼º ¹× ½Åȣó¸® ÇÐȸ·Î ÀÎÁ¤¹Þ°í ÀÖ´Ù.

³×À̹ö¿Í ¶óÀÎÀº ¿ÃÇØ 6¿ù °³ÃÖµÈ ‘ICASSP 2021’¿¡¼­µµ 14°ÇÀÇ ³í¹®À» ¹ßÇ¥Çß´Ù.

'ÀÎÅͽºÇÇÄ¡ 2021' äÅÃÇÑ ³×À̹ö ³í¹® ¸®½ºÆ®
1. High-fidelity Parallel WaveGAN with Multi-band Harmonic-plus-Noise Model
- Ȳ¹ÎÁ¦, Ryuichi Yamamoto (LINE), ¼ÛÀº¿ì, ±èÀç¹Î
- À½¼º ¸ðµ¨¸µ ±â¹ýÁß ÇϳªÀÎ multi-band harmonic-plus-noise ¸ðµ¨À» Parallel WaveGAN ¿¡ Àû¿ëÇØ º¸ÄÚ´õÀÇ Ç°Áú°ú ¾ÈÁ¤¼ºÀ» Çâ»ó.
 
2. LiteTTS: A Decoder-free Light-weight Text-to-wave Synthesizer Based on Generative Adversarial Networks
- Kim Nguyen (¿¬¼¼´ë), Á¤±âÇõ (¿¬¼¼´ë), ¾ö¼¼¿¬ (¿¬¼¼´ë), Ȳ¹ÎÁ¦, ¼ÛÀº¿ì, °­È«±¸ (¿¬¼¼´ë)
- Text-to-feature ¿Í feature-to-wave ·Î ºÐ¸®µÇ¾îÀÖ´Â TTS ÆÄÀÌÇÁ¶óÀÎÀ» Çϳª·Î ¹­´Â text-to-wave ¸ðµ¨À» Á¦¾È.
 
3. Label Embedding for Chinese Grapheme-to-Phoneme Conversion
- ÃÖÀººñ (KAIST), ±èÈ­¿¬, ±èÁ¾È¯, ±èÀç¹Î
- Chinese polyphone conversion problemÀ» À§ÇÑ label embedding approach¸¦ Á¦¾È.
 
4. Look Who’s Talking: Active Speaker Detection in the Wild
- ±èÀ¯Áø, ÇãÈñ¼ö, ÃÖ¼Ò¿¬, Á¤¼öȯ, ±ÇÀ¯È¯, À̺ÀÁø, ±Ç¿µ±â, Á¤Áؼ±
- ¸ÖƼ¸ð´Þ Active Speaker Detection µ¥ÀÌÅͼ Á¦ÀÛ °ø°³
 
5. Adapting Speaker Embeddings for Speaker Diarisation
- ±Ç¿µ±â, Á¤Áö¿ø, ÇãÈñ¼ö, ±èÀ¯Áø, À̺ÀÁø, Á¤Áؼ±
- Speaker Diarization ¼º´É °³¼±À» À§ÇÑ Speaker Embedding °­È­ ¹æ¹ý ¿¬±¸
 
6. Three-class Overlapped Speech Detection using a Convolutional Recurrent Neural Network
- Á¤Áö¿ø, ÇãÈñ¼ö, ±Ç¿µ±â, Á¤Áؼ±, À̺ÀÁø
- 3°³ Ŭ·¡½º+CRNNÀ» ÀÌ¿ëÇÑ Overlapped Speech Detection ±â¹ý Á¦¾È
 
7. Graph Attention Networks for Anti-Spoofing
- Hemlata Tak (EURECOM), Á¤Áö¿ø, Jose Patino (EURECOM), Massimiliano Todisco (EURECOM), Nicholas Evans (EURECOM)
- GNNÀ» ÀÌ¿ëÇÑ audio spoofing detection ¿¬±¸
 
8. DEMUCS-Mobile : On-device lightweight speech enhancement
- ÀÌ·çÄ«½º, ÁöÀ¯³ª, À̹ÎÀç, Ãֹμ®
- CNN µö·¯´× ±â¹Ý ÀâÀ½ Á¦°Å ¸ðµ¨ÀÇ °æ·®È­ ±â¹ýÀ» Á¦¾ÈÇÏ¿© ¸ð¹ÙÀÏ on-device¿¡¼­ ½Ç½Ã°£À¸·Î µ¿ÀÛ°¡´ÉÇÑ ÀâÀ½Á¦°Å ¸ðµ¨ ±¸Çö
 
9. Layer Pruning on Demand with Intermediate CTC
- ÀÌÀç¼Û, °­Áø±¸, Shinji Watanabe (CMU)
- Transformer+CTC ¸ðµ¨À» fine-tuning ¾øÀÌ pruningÇϱâ À§ÇÑ ÇнÀ ¹æ¹ý·Ð

‘ÀÎÅͽºÇÇÄ¡ 2021’¿¡¼­ äÅÃµÈ ³×À̹öÀÇ ¿¬±¸ ³í¹®Àº À½¼ºÀνÄ, À½¼ºÇÕ¼º, µ¥ÀÌÅͼ Á¦ÀÛ µî À½¼º ¹× ½ÅÈ£¿Í °ü·ÃµÈ ´Ù¾çÇÑ ºÐ¾ß¸¦ Æ÷°ýÇÑ´Ù.

ÀϺΠ³í¹®Àº ³×À̹öÀÇ ½ÇÁ¦ ¼­ºñ½º¿¡ Àû¿ëµÇ¸ç »ç¿ëÀÚ¿¡°Ô ´õ¿í Æí¸®ÇÑ ¼­ºñ½º °æÇèÀ» Á¦°øÇÏ°í ÀÖ´Ù.

À½¼º ÇÕ¼º Ç°Áú Çâ»ó ±â¼ú ¿¬±¸´Â ‘Ŭ·Î¹Ù´õºù’, ‘Ŭ·Î¹Ù ½º¸¶Æ® ½ºÇÇÄ¿’, ‘³×À̹ö AiCALL’ µî ³×À̹ö Ŭ·Î¹Ù¿¡¼­ Á¦°øÇÏ´Â ´Ù¾çÇÑ À½¼º ÇÕ¼º ¼­ºñ½º¿¡ Àû¿ëµÆ´Ù.

¿©·¯ È­ÀÚ°¡ µ¿½Ã¿¡ ¸»ÇÏ´Â »óȲ¿¡¼­ ‘È­ÀÚ ºÐÇÒ(Speaker Diarisation)’À» À§ÇØ °¢ ¹ßÈ­ÀÚÀÇ Æ¯Â¡À» ´õ¿í Á¤È®ÇÏ°Ô ÇнÀÇÏ´Â ±â¹ý¿¡ ´ëÇÑ ¿¬±¸´Â ‘Ŭ·Î¹Ù³ëÆ®’ ¼­ºñ½º °³¼± µî¿¡ È°¿ëµÇ°í ÀÖ´Ù.

¶ÇÇÑ ³×À̹ö°¡ ¹ßÇ¥ÇÑ 9°ÇÀÇ ³í¹® Áß 5°ÇÀº ÄÄÇ»ÅÍ °úÇÐ ¹× Á¤º¸ ½Ã½ºÅÛ ºÐ¾ß¿¡¼­ ¼±µÎÁÖÀÚ·Î ²ÅÈ÷´Â ÇÁ¶û½º ¿¬±¸±â°ü À¯·¹ÄÄ(EURECOM)À» ºñ·ÔÇØ, Ä«³×±â¸á·Ð´ëÇб³, Ä«À̽ºÆ®(KAIST), ¿¬¼¼´ëÇб³, ¶óÀÎ µî AI ±â¼úÀ» À̲ô´Â ±¹³»¿Ü ±â¾÷ ¹× ¿¬±¸ ±â°üµé°ú Çù·Â ¿¬±¸·Î ¼öÇàµÆ´Ù.

³×À̹ö´Â ±Û·Î¹ú AI ¿¬±¸Çõ½Å »ýÅÂ°è ±¸ÃàÀ» À§ÇÑ ³ë·ÂÀÌ °¡½ÃÀûÀÎ ¼º°ú·Î µå·¯³ª°í ÀÖ´Ù°í µ¡ºÙ¿´´Ù.

À̹ø ¼º°ú¸¦ ±â¹ÝÀ¸·Î ³×À̹ö´Â ±Û·Î¹ú AI R&D »ýÅ°踦 Áö¼ÓÀûÀ¸·Î È®ÀåÇØ°¡°Ú´Ù´Â °èȹÀÌ´Ù. 

ÇöÀç ¶óÀΰúÀÇ ¿¬°è¸¦ Áß½ÉÀ¸·Î ÇÑ ÀϺ»À» ³Ñ¾î, º£Æ®³²¿¡¼­´Â Ç㽺Ʈ(HUST), PTIT¿Í °øµ¿¿¬±¸¼¾Å͸¦ ¼³¸³ÇßÀ¸¸ç, À¯·´¿¡¼­´Â ÇÁ¶û½ºÀÇ ³×À̹ö·¦½ºÀ¯·´°ú ¿¬±¸Çù·ÂÀ» À̾°í ÀÖ´Ù.

Áö³­ 7¿ù¿¡´Â Àΰø½Å°æ¸Á°ú ·Îº¿°øÇÐ ºÐ¾ß µ¶ÀÏÀÇ Æ¢ºù°Õ´ëÇб³¿Í °øµ¿¿¬±¸¼¾ÅÍ ¼³¸³¿¡ ÇÕÀÇÇϱ⵵ Çß´Ù.

Á¤¼®±Ù ³×À̹ö Ŭ·Î¹Ù CIC ´ëÇ¥´Â “R&D¿¡ ´ëÇÑ Àû±ØÀûÀÎ ÅõÀÚ¸¦ ¹ÙÅÁÀ¸·Î ³×À̹öÀÇ AI ¿¬±¸Çõ½Å »ýÅ°è´Â ±× ±Ô¸ð¸¦ ´õ¿í Å°¿ö°¡°í ÀÖ´Ù”¸ç “ÀÎÅͽºÇÇÄ¡ 2021¿¡¼­ ³×À̹ö¿Í ¶óÀÎ, ±×¸®°í ±Û·Î¹ú AI ¿¬±¸ÁøÀÌ ¶Ù¾î³­ ¼º°ú¸¦ °ÅµÐ °Íó·³, ¾ÕÀ¸·Î ³×À̹ö¸¦ Áß½ÉÀ¸·Î ±¸ÃàµÉ ´õ¿í ´Ù¾çÇÑ AI R&D »ýÅ°迡¼­ »õ·Î¿î ¼º°ú°¡ ³ª¿Ã ¼ö ÀÖÀ» °ÍÀ¸·Î ±â´ëÇÑ´Ù”°í ¸»Á™´Ù.

<±èµ¿±â ±âÀÚ>kdk@bikorea.net

< ÀúÀÛ±ÇÀÚ © BI KOREA ¹«´ÜÀüÀç ¹× Àç¹èÆ÷±ÝÁö >
±èµ¿±â ±âÀÚÀÇ ´Ù¸¥±â»ç º¸±â  
ÆùƮŰ¿ì±â ÆùÆ®ÁÙÀ̱â ÇÁ¸°Æ®Çϱ⠸ÞÀϺ¸³»±â ½Å°íÇϱâ
Æ®À§ÅÍ ÆäÀ̽ººÏ ¹ÌÅõµ¥ÀÌ ¿äÁò ³×À̹ö ±¸±Û msn µÚ·Î°¡±â À§·Î°¡±â
ÀÌ ±â»ç¿¡ ´ëÇÑ ´ñ±Û À̾߱â (0)
ÀÚµ¿µî·Ï¹æÁö¿ë Äڵ带 ÀÔ·ÂÇϼ¼¿ä!   
È®ÀÎ
- 200ÀÚ±îÁö ¾²½Ç ¼ö ÀÖ½À´Ï´Ù. (ÇöÀç 0 byte / ÃÖ´ë 400byte)
- ¿å¼³µî ÀνŰø°Ý¼º ±ÛÀº »èÁ¦ ÇÕ´Ï´Ù. [¿î¿µ¿øÄ¢]
ÀÌ ±â»ç¿¡ ´ëÇÑ ´ñ±Û À̾߱â (0)
1
Èï±¹»ý¸í, 800¾ï ±Ô¸ð ¡®±â°£°è ±¸Ã࡯ ÀÔÂû °ø°í
2
<ÃÊÁ¡>Èï±¹»ý¸í ¡®±â°£°è ±¸Ã࡯ RFP¸¦ ºÐ¼®ÇØ º¸´Ï¡¦
3
´õÁ¸ºñÁî¿Â-¼¼Á¤±×·ì, ¡®Â÷¼¼´ë Á¤º¸½Ã½ºÅÛ¡¯ ±¸Ãà
4
½´ÆÛ¸¶ÀÌÅ©·Î, ¿£ºñµð¾Æ ±â¹Ý ¡®»ý¼ºÇü AI ½´ÆÛŬ·¯½ºÅÍ¡¯ 3Á¾ Ãâ½Ã
5
µ¨ Å×Å©³î·ÎÁö½º, ¿£ºñµð¾Æ¡¤½ºÅ¸¹ö½ºÆ® µî Çù·Â È®´ë
6
KT, ÆDZ³ ½Å»ç¿Á¿¡ ¡®¿ÀÇ À̳뺣ÀÌ¼Ç ¼¾ÅÍ¡¯ °³¼Ò
7
ÄÚ½ºÄÞ, ¡°Àå¾ÖÀÎ IT â¾÷ µ½´Â´Ù¡±
8
¿ì¸®ÀºÇà, °ü¾Ç±¸ 21°³ ÀüÅë½ÃÀå¿¡ °áÁ¦ ´Ü¸»±â ¹«»ó ±³Ã¼
9
¾È·¦, Á¤ºÎ SaaS À°¼º¿¡ ¡®MSP Áö¿ø±â¾÷¡¯À¸·Î ¼±Á¤
10
¡®ÇϳªÀºÇà ºÎ´ÙÆ佺Ʈ »ç¹«¼Ò¡¯ °³¼Ò
ȸ»ç¼Ò°³¤ý±â»çÁ¦º¸¤ý±¤°í¹®ÀǤýºÒÆí½Å°í¤ý°³ÀÎÁ¤º¸Ãë±Þ¹æħ¤ýû¼Ò³âº¸È£Á¤Ã¥¤ýÀ̸ÞÀϹ«´Ü¼öÁý°ÅºÎ
¼­¿ï½Ã ¿µµîÆ÷±¸ ¿©ÀÇ´ë¹æ·Î65±æ 13, 904(¿©Àǵµµ¿ À¯Ã¢ºôµù)| Tel: 02-785-5108 | Fax 02-785-5109
À̸ÞÀÏÁÖ¼Ò¹«´Ü¼öÁý°ÅºÎ | (ÁÖ)ºñ¾ÆÀÌÄÚ¸®¾Æ´å³Ý | ´ëÇ¥ÀÌ»ç : ±èµ¿±â | »ç¾÷ÀÚ µî·Ï¹øÈ£:107-87-99085 | °³ÀÎÁ¤º¸°ü¸®Ã¥ÀÓÀÚ : ±èµ¿±â
µî·Ï¹øÈ£ : ¼­¿ï ¾Æ01269 | µî·ÏÀÏÀÚ 2010.06.18, ½Å°íÀÏÀÚ 2008.10.22 | ¹ßÇàÀÎ:±èµ¿±â | ¹ßÇàÀÏÀÚ:2010.06.01 | ÆíÁýÀÎ : ±èµ¿±â
û¼Ò³âº¸È£Ã¥ÀÓÀÚ : ±èµ¿±â
Copyright © 2012 BI KOREA. All rights reserved. mail to webmaster@bikorea.net