PaddlePaddle 2.0 is now available! The latest version of our #DeepLearning platform features dynamic computational graphs, a new API system, distributed training for trillion-parameter models, and better hardware support. Read more below.
Baidu Research
1,822 posts
Baidu Research
@BaiduResearch
Baidu Research brings together top talent from around the world to focus on future-looking fundamental researches in #AI #deeplearning #machinelearning.
- Understanding the future of #tech is more important than ever. That's why we've released our prediction for the top 10 tech trends to watch in 2021, summarized below in a video synthesized by our #AI. See the full scope of the trends here: research.baidu.com/Blog/index-vie…
00:00 - Registration for the 2021 Language & Intelligence Challenge is now open! This year's competition is based on LUGE, an open-source project of Chinese NLP benchmarks. Winners will share an RMB 300,000 prize pool, register by May 12th to enter! Learn more: lic2021.ccf.org.cn
- Time for another benchmark competition! We're excited to co-organize the Neural Architecture Search workshop at #CVPR2021. There will be 3 competition tracks: supernet, performance estimation & unseen data—submit by May 17 to enter the $10,000 prize pool. cvpr21-nas.com/competition
- Our neural network based system learned to "clone" a voice with less than a minute of audio data from the speaker. Check out our paper to find out more about this latest breakthrough in speech synthesis. #DeepLearning #MachineLearning #AI bit.ly/2GvGhBP
- Calling all researchers, academics, and colleagues in the fields of NLP & document image understanding: we're seeking papers for our Workshop on Document Image & Language at the ICDAR 2021 in Lausanne, CH. Best paper receives a $1000 reward! Learn more: dil2021.github.io/index.html
- Introducing ERNIE-ViLG 2.0, a 24B text-to-image model with knowledge-enhanced Mixture-of-Denoising-Experts that obtains SOTA results on MS-COCO. This is also the largest text-to-image model at present. Paper: arxiv.org/pdf/2210.15257… Demo @huggingface: huggingface.co/spaces/PaddleP…
- Announcing the new #PaddlePaddle at the Wave Summit+ 2019 in Beijing today! Baidu #DL platform brings 21 significant new features (100+ models, Paddle Lite 2.0, graph & federated learning frameworks) to help developers build industry-grade #AI models. github.com/PaddlePaddle/P…
- Introducing PP-YOLO, a new implementation of object detector based on PaddlePaddle. Compared with YOLOv4, PP-YOLO improves mAP on COCO to 45.9% and runs at 72.9 fps on Tesla V100. #ComputerVision #AI Paper: arxiv.org/abs/2007.12099 GitHub: github.com/PaddlePaddle/P…
00:00 - Our team has reached a new milestone in speech synthesis with the release of ClariNet, the first fully end-to-end TTS model that directly converts text into waveform with a single neural network. It generates all samples of audio waveform in parallel. #AI bit.ly/2JCuJ0Y
00:00 - Just released! The Docker-based Jupyter Notebook version of PaddlePaddle's Deep Learning 101 Book. bit.ly/2mN9FtT
- Learn #deeplearning from this tutorial ranked as the No.1#MachineLearning article for the past year by @Mybridge bit.ly/2muJGK6
- Proud to announce our paper "Deep Voice 2: Multi-Speaker Neural Text-to-Speech" is accepted for #nips2017 spotlight! bit.ly/2rUG7xl
- Our team is open sourcing NCRF, a #DeepLearning based algorithm for more accurate detection of tumor cells. Read more about it here: bit.ly/2JM1DBk #AI #MachineLearning
