Abstract | ||
---|---|---|
Depth information plays an important role in the human visual system, however it is not yet well- explored in existing proposal generation models. In this paper we propose a new geocentric embedding for depth images that encodes depth to the camera, the structural edges and height above ground-plane for each pixel named DEH channels. We demonstrate that this geocentric embedding works can be use to generate high quality object proposals with convolutional neural networks. Our experiments show significant performance improvements over existing RGB and RGB-D object proposal methods on the challenging KITTI benchmark. We also exploit Fast R- CNN [8] on top of these proposals to perform object detection with RGB images, our approach obtains state-of-the-art results on all three KITTI object classes. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1109/SMARTCOMP.2017.7947019 | 2017 IEEE International Conference on Smart Computing (SMARTCOMP) |
Keywords | Field | DocType |
generic feature learning,object proposals generation,geocentric embedding,depth images,DEH channels,convolutional neural networks,RGB-D object proposal methods,KITTI benchmark,object detection,fast R-CNN | Computer vision,Object detection,Embedding,Convolutional neural network,Computer science,Human visual system model,Communication channel,RGB color model,Artificial intelligence,Pixel,Benchmark (computing) | Conference |
ISBN | Citations | PageRank |
978-1-5090-6518-9 | 0 | 0.34 |
References | Authors | |
16 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Yiliu Feng | 1 | 6 | 3.13 |
Zhengfa Liang | 2 | 27 | 7.58 |
Hengzhu Liu | 3 | 86 | 23.28 |