ニュース&特集
読み込み中…
Microsoft Research ブログ
HEXA: Self-supervised pretraining with hard examples improves visual representations
| Chunyuan Li, Lei Zhang, と Jianfeng Gao
Humans perceive the world through observ…
Microsoft Research ブログ
VinVL: Advancing the state of the art for vision-language models
| Pengchuan Zhang, Lei Zhang, と Jianfeng Gao
Humans understand the world by perceivin…
Microsoft Research ブログ
A holistic representation toward integrative AI
| Xuedong Huang
At Microsoft, we have been on a quest to…
Microsoft Research ブログ
Novel object captioning surpasses human performance on benchmarks
| Kevin Lin, Xiaowei Hu, と Lijuan Wang
Consider for a moment what it takes to v…
Microsoft Research ブログ
Enhancing your photos through artificial intelligence
The amount of visual data we accumulate …
Microsoft Research ブログ
Objects are the secret key to revealing the world between vision and language
| Chunyuan Li, Lei Zhang, と Jianfeng Gao
Humans perceive the world through many c…
Microsoft Research ブログ
Expanding scene and language understanding with large-scale pre-training and a unified architecture
| Hamid Palangi
Making sense of the world around us is a…
Microsoft Research ブログ
A picture from a dozen words – A drawing bot for realizing everyday scenes—and even stories
If you were asked to draw a picture of s…
Microsoft Research ブログ
Microsoft Pix’s AI gets even smarter, Business Card feature works with LinkedIn to make it easier than ever to manage your contacts
Today’s update from Microsoft debuts a n…