Towards Cost-Effective and Performance-Aware Vision Algorithms

@inproceedings{Dai2016TowardsCA,
  title={Towards Cost-Effective and Performance-Aware Vision Algorithms},
  author={Dengxin Dai},
  year={2016}
}
Computer vision has leaped forward during the last decade, and now is able to recognize objects of thousands of categories and reconstruct 3D scenes at cityor world-scale. However, the field still has to find means to keep up with the exploration of the massive amounts of data being captured on a daily basis. This is mainly due to the lack of sufficient training annotations and the lack of computational resources. The thesis is dedicated to mitigate the problem. Firstly, we elaborate two… CONTINUE READING

Citations

Publications citing this paper.
SHOWING 1-4 OF 4 CITATIONS

Efficient Object Annotation via Speaking and Pointing

VIEW 1 EXCERPT
CITES BACKGROUND

Failure Prediction for Autonomous Driving

VIEW 1 EXCERPT
CITES BACKGROUND

Object Referring in Visual Scene with Spoken Language

VIEW 1 EXCERPT
CITES BACKGROUND