DeepID-Net: multi-stage and deformable deep convolutional neural networks for object detection.
Wanli OuyangPing LuoXingyu ZengShi QiuYonglong TianHongsheng LiShuo YangZhe WangYuanjun XiongChen QianZhenyao ZhuRuohui WangChen Change LoyXiaogang WangXiaoou TangPublished in: CoRR (2014)
Keyphrases
- multistage
- convolutional neural networks
- object detection
- deformable part models
- convolutional network
- dynamic programming
- single stage
- stochastic programming
- production system
- computer vision
- lot sizing
- stochastic optimization
- object categories
- coarse to fine
- multi class
- object recognition
- deformable models
- object detectors
- optimal policy
- attack detection
- assembly systems
- action detection
- lot streaming
- production line
- action recognition
- search algorithm