Beijing Megvii Co., Ltd.旷视的核心技术是计算视觉及传感技术相关的人工智能算法,包括但不限于人脸识别、人体识别、手势识别、文字识别、证件识别、图像识别、物体识别、车牌识别、视频分析、三维重建、智能传感与控制等技术。旷视通过底层AI算法引擎和AIoT操作系统的建设实现技术商业化。
computer vision and pattern recognition, (2019)
The uncertainties in large-scale object detection datasets can hinder the performance of state-of-theart object detectors
Cited by89BibtexViews126
0
0
CVPR, pp.9336-9345, (2019)
We propose a novel Progressive Scale Expansion Network to successfully detect the text instances with arbitrary shapes in the natural scene images
Cited by84BibtexViews43
0
0
CVPR, (2019)
For the proposed MLGCN, we report the results based on the binary correlation matrix and the re-weighted correlation matrix, respectively
Cited by68BibtexViews51Links
0
0
CVPR, (2019): 3313-3322
We propose an end-to-end neural network for depth prediction from Sparse LiDAR data and a single color image
Cited by60BibtexViews56Links
0
0
CVPR, (2019): 9522-9531
We propose deep feature aggregation to tackle real-time semantic segmentation on high resolution image
Cited by55BibtexViews10Links
0
0
AAAI, (2019)
We have presented an end-to-end framework for video inpainting through a joint 2D-3D CNN which contains a temporal structure inference network and spatial detail recovering network
Cited by48BibtexViews40Links
0
0
national conference on artificial intelligence, (2019)
We have presented a shape robust text detector that can detect text with arbitrary shapes
Cited by46BibtexViews47Links
0
0
Minghui Liao, Jian Zhang,Zhaoyi Wan, Fengming Xie, Jiajun Liang, Pengyuan Lyu,Cong Yao,Xiang Bai
national conference on artificial intelligence, (2019)
We have presented a method called Character Attention FCN for scene text recognition, which models the problem in a two-dimensional fashion
Cited by42BibtexViews36Links
0
0
CVPR, (2019): 6172-6181
It is easy to find that Segmentation Quality is the common mean IOU metric normalized for matching instances, and Detection Quality could be regarded as a form of detection accuracy
Cited by41BibtexViews57Links
0
0
CVPR, (2019): 1575-1584
We propose a novel upscale module named MetaUpscale to solve the super-resolution of arbitrary scale factor with a single model
Cited by40BibtexViews54Links
0
0
CVPR, (2019)
Experimental results confirm that Visibility-aware Part Model surpasses both the global feature learning baseline and part-based convolutional methods, and the achieved performance is on par with the state of the art
Cited by34BibtexViews39Links
0
0
International Conference on Computer Vision, (2019): 3296-3305
We have presented MetaPruning for channel pruning with following advantages: 1) it achieves much higher accuracy than the uniform pruning baselines as well as other state-of-the-art channel pruning methods, both traditional and AutoML-based; 2) it can flexibly optimize with respe...
Cited by23BibtexViews144Links
0
0
NeurIPS, pp.6638-6648, (2019)
As in Fig. 1, DetNAS consists of 3 steps: supernet pre-training on ImageNet, supernet fine-tuning on detection datasets and architecture search on the trained supernet
Cited by21BibtexViews122
0
0
CVPR, (2019): 998-1008
We address the problem of modeling local geometric structure amongst points with geometric structure by applying a convolutional-like operation operation and a hierarchical feature extraction framework dubbed Geo-convolutional neural networks
Cited by21BibtexViews56Links
0
0
Tao Hu, Pengwan Yang, Chiliang Zhang,Gang Yu,Yadong Mu,Cees Snoek
national conference on artificial intelligence, (2019)
Context features in equal level are fused by our MCG module, which highly facilitates the support branch to globally “support” the query branch
Cited by18BibtexViews68Links
0
0
Mingkun Yang, Yushuo Guan, Minghui Liao, Xin He,Kaigui Bian,Song Bai,Cong Yao,Xiang Bai
international conference on computer vision, pp.9147-9156, (2019)
We have proposed a Symmetry-constrained Rectification Network for scene text recognition
Cited by17BibtexViews18Links
0
0
CoRR, (2019): 8709-8718
We build a differentiable neural renderer to render the strokes, which allows using model-based Deep Reinforcement Learning algorithms to further improve the quality of recreated images
Cited by14BibtexViews7Links
0
0
Ruihang Chu,Yifan Sun, Yadong Li, Zheng Liu, Chi Zhang,Yichen Wei
ICCV, pp.8281-8290, (2019)
Experimental results confirm that viewpoint-aware network significantly improves re-ID accuracy, especially when the positive pairs are observed from different viewpoints and the negative pairs are observed from similar viewpoints
Cited by13BibtexViews13Links
0
0
Wenhai Wang,Enze Xie, Xiaoge Song, Yuhang Zang, Wenjia Wang, Tong Lu,Gang Yu,Chunhua Shen
2991626090, pp.8440-8449, (2019)
With SynthText pre-training, the F-measure of Pixel Aggregation Network-320 boosting to 79.9%, and the best F-measure achieve by PAN-640 is 85.0%, which is 2.1% better than second-best SPCNet
Cited by13BibtexViews41
0
0
CVPR, (2019): 1419-1428
This paper presents GIF2Video, the first learning-based method for enhancing the visual quality of Graphics Interchange Format in the wild
Cited by11BibtexViews29Links
0
0
Keywords
Face Recognition300-w Facial Landmark LocalizationCoarse-to-fine Convolutional Network CascadeConvolutional Neural NetworkCurrent Network LevelDeep Convolutional Neural NetworkDeep LearningExtensive Facial LandmarkExtensive Facial Landmark LocalizationFace Search
Authors
Jian Sun
Paper 19
Gang Yu
Paper 13
Xiangyu Zhang
Paper 11
Yuning Jiang
Paper 10
Zhimin Cao
Paper 8
Jue Wang
Paper 8
Haoqiang Fan
Paper 6
Cong Yao
Paper 6
Chao Peng
Paper 6
Tete Xiao
Paper 5