Sign In

JoyAI-Image: Awaking Spatial Intelligence in Unified Multimodal Understanding and Generation

์ž‘์„ฑ์ž
  • Haebom
์นดํ…Œ๊ณ ๋ฆฌ
Empty

์ €์ž

Lin Song, Wenbo Li, Guoqing Ma, Wei Tang, Bo Wang, Yuan Zhang, Yijun Yang, Yicheng Xiao, Jianhui Liu, Yanbing Zhang, Guohui Zhang, Wenhu Zhang, Hang Xu, Nan Jiang, Xin Han, Haoze Sun, Maoquan Zhang, Haoyang Huang, Nan Duan

๐Ÿ’ก ๊ฐœ์š”

๋ณธ ๋…ผ๋ฌธ์€ ์‹œ๊ฐ์  ์ดํ•ด, ํ…์ŠคํŠธ-์ด๋ฏธ์ง€ ์ƒ์„ฑ, ์ง€์‹œ ๊ธฐ๋ฐ˜ ์ด๋ฏธ์ง€ ํŽธ์ง‘์„ ํ†ตํ•ฉํ•˜๋Š” ๋ฉ€ํ‹ฐ๋ชจ๋‹ฌ ๊ธฐ์ดˆ ๋ชจ๋ธ์ธ JoyAI-Image๋ฅผ ์ œ์•ˆํ•ฉ๋‹ˆ๋‹ค. JoyAI-Image๋Š” ๊ณต๊ฐ„ ์ธ์ง€ ๋Šฅ๋ ฅ์ด ๊ฐ•ํ™”๋œ MLLM๊ณผ MMDiT๋ฅผ ๊ฒฐํ•ฉํ•˜์—ฌ ์ง€๊ฐ๊ณผ ์ƒ์„ฑ์ด ๊ณต์œ  ๋ฉ€ํ‹ฐ๋ชจ๋‹ฌ ์ธํ„ฐํŽ˜์ด์Šค๋ฅผ ํ†ตํ•ด ์ƒํ˜ธ์ž‘์šฉํ•˜๋„๋ก ํ•ฉ๋‹ˆ๋‹ค. ์ด ๋ชจ๋ธ์€ ํ†ตํ•ฉ๋œ ์ง€์‹œ ํŠœ๋‹, ๊ธด ํ…์ŠคํŠธ ๋ Œ๋”๋ง ๊ฐ๋…, ๊ณต๊ฐ„์ ์œผ๋กœ ๊ธฐ๋ฐ˜ํ•œ ๋ฐ์ดํ„ฐ, ์ผ๋ฐ˜ ๋ฐ ๊ณต๊ฐ„ ํŽธ์ง‘ ์‹ ํ˜ธ๋ฅผ ํ™œ์šฉํ•˜์—ฌ ์ผ๋ฐ˜์ ์ธ ๋ฉ€ํ‹ฐ๋ชจ๋‹ฌ ๋Šฅ๋ ฅ๋ฟ๋งŒ ์•„๋‹ˆ๋ผ ๊ธฐํ•˜ํ•™์  ์ถ”๋ก  ๋ฐ ์ œ์–ด ๊ฐ€๋Šฅํ•œ ์‹œ๊ฐ ํ•ฉ์„ฑ ๋Šฅ๋ ฅ์„ ๊ฐ•ํ™”ํ•ฉ๋‹ˆ๋‹ค.

๐Ÿ”‘ ์‹œ์‚ฌ์  ๋ฐ ํ•œ๊ณ„

โ€ข
JoyAI-Image๋Š” ์‹œ๊ฐ ์ดํ•ด, ํ…์ŠคํŠธ-์ด๋ฏธ์ง€ ์ƒ์„ฑ, ์ง€์‹œ ๊ธฐ๋ฐ˜ ์ด๋ฏธ์ง€ ํŽธ์ง‘ ๋“ฑ ๋‹ค์–‘ํ•œ ๋ฉ€ํ‹ฐ๋ชจ๋‹ฌ ํƒœ์Šคํฌ์—์„œ ์ตœ์ฒจ๋‹จ ๋˜๋Š” ๊ฒฝ์Ÿ๋ ฅ ์žˆ๋Š” ์„ฑ๋Šฅ์„ ๋‹ฌ์„ฑํ•˜๋ฉฐ, ํŠนํžˆ ๊ณต๊ฐ„ ์ง€๋Šฅ์„ ๊ฐ•ํ™”ํ•ฉ๋‹ˆ๋‹ค.
โ€ข
ํ–ฅ์ƒ๋œ ์ดํ•ด, ์ œ์–ด ๊ฐ€๋Šฅํ•œ ๊ณต๊ฐ„ ํŽธ์ง‘, ์ƒˆ๋กœ์šด ์‹œ์  ๋ณด์กฐ ์ถ”๋ก  ๊ฐ„์˜ ์–‘๋ฐฉํ–ฅ ์ƒํ˜ธ์ž‘์šฉ์€ ๋ชจ๋ธ์ด ์ผ๋ฐ˜์ ์ธ ์‹œ๊ฐ ๋Šฅ๋ ฅ์„ ๋„˜์–ด ๊ฐ•๋ ฅํ•œ ๊ณต๊ฐ„ ์ง€๋Šฅ์œผ๋กœ ๋‚˜์•„๊ฐˆ ์ˆ˜ ์žˆ์Œ์„ ๋ณด์—ฌ์ค๋‹ˆ๋‹ค.
โ€ข
์ด๋Š” ๋น„์ „-์–ธ์–ด-ํ–‰๋™ ์‹œ์Šคํ…œ ๋ฐ ์›”๋“œ ๋ชจ๋ธ๊ณผ ๊ฐ™์€ ๋‹ค์šด์ŠคํŠธ๋ฆผ ์• ํ”Œ๋ฆฌ์ผ€์ด์…˜์—์„œ ํ†ตํ•ฉ ์‹œ๊ฐ ๋ชจ๋ธ์„ ์œ„ํ•œ ์œ ๋งํ•œ ๊ฒฝ๋กœ๋ฅผ ์ œ์‹œํ•ฉ๋‹ˆ๋‹ค.
โ€ข
๋ณธ ์—ฐ๊ตฌ๋Š” ๋ชจ๋ธ์˜ ํ™•์žฅ์„ฑ๊ณผ ๋‹ค์–‘ํ•œ ๋ฐ์ดํ„ฐ์…‹์— ๋Œ€ํ•œ ์ผ๋ฐ˜ํ™” ์„ฑ๋Šฅ์„ ๋”์šฑ ํ–ฅ์ƒ์‹œํ‚ค๊ธฐ ์œ„ํ•œ ์ถ”๊ฐ€์ ์ธ ์—ฐ๊ตฌ๊ฐ€ ํ•„์š”ํ•ฉ๋‹ˆ๋‹ค.
๐Ÿ‘
Made with Slashpage