“Background replacement” by Penta
Conference:
Type(s):
Title:
- Background replacement
Presenter(s)/Author(s):
Abstract:
Researchers have proposed various techniques for automatically creating composite images [Lalonde et al. 2007; Hays and Efros 2007]. In this paper, we present a technique for automatically replacing the background of an image with a better background image. User specifies the type of background he/she wants in the new composite. Type of the background can be ”beach”, ”landscape”,”lake” or ”mountains”, etc. For each background category, a large dataset of background images are collected based on the tags from the web. Datasets collected in this fashion can have a lot of irrelevant images. We manually choose some candidate backgrounds for each category as representatives. A gist descriptor [Oliva and Torralba 2006] is computed for all the images. We retain the ones that are close to these candidate background images with respect to their gist scene descriptor.
References:
1. Hays, J. H., and Efros, A. A. 2007. Scene completion using millions of photographs. ACM Transactions on Graphics (SIGGRAPH 2007) 26, 3 (August).
2. Hoiem, D., Efros, A. A., and Hebert, M. 2005. Geometric context from a single image. In International Conference of Computer Vision (ICCV), IEEE, vol. 1, 654–661.
3. Lalonde, J.-F., Hoiem, D., Efros, A. A., Rother, C., Winn, J., and Criminisi, A. 2007. Photo clip art. ACM Transactions on Graphics (SIGGRAPH 2007) 26, 3 (August).
4. Oliva, A., and Torralba, A. 2006. Building the gist of a scene: The role of global image features in recognition. Visual Perception, Progress in Brain Research 155.
5. Wang, J., and Cohen, M. 2007. Optimized color sampling for robust matting. In IEEE Conference on Computer Vision & Pattern Recognition.
