Diem Vu

Kernelized structural SVM learning for supervised object segmentation

Object segmentation needs to be driven by top-down knowledge to produce semantically meaningful results. In this paper, we propose a supervised segmentation approach that tightly integrates object-level top down information with low-level image cues. The information from the two levels is fused under a kernelized structural SVM learning framework. We defined a novel nonlinear kernel for comparing two image-segmentation masks. This kernel combines four different kernels: the object similarity kernel, the object shape kernel, the per-image color distribution kernel, and the global color distribution kernel. Our experiments show that the structured SVM algorithm finds bad segmentations of the training examples given the current scoring function and punishes these bad segmentations to lower scores than the example (good) segmentations. The result is a segmentation algorithm that not only knows what good segmentations are, but also learns potential segmentation mistakes and tries to avoid them. Our proposed approach can obtain comparable performance to other state-of-the-art top-down driven segmentation approaches yet is flexible enough to be applied to widely different domains.

Visual search engine for product images

Nowadays there are many product comparison web sites. But most of them only use text information. This paper introduces a novel visual search engine for product images, which provides a brand-new way of visually locating products through Content-based Image Retrieval (CBIR) technology. We discusses the unique technical challenges, solutions, and experimental results in the design and implementation of this system.

An efficient fashion-driven learning approach to model user preferences in on-line shopping scenarios

In this work we tackle the problem of search personalization for on-line soft goods shopping. By learning what the user likes and what the user does not like, better search rankings and therefore a better overall shopping experience can be obtained. The first contribution of the work is in terms of feature selection: given the specific nature of the domain, we combine the traditional visual and text feature into a fashion-driven low dimensional space, compact yet very discriminative. On the learning stage, we describe a two step hybrid learning algorithm, that combines a discriminative model learned off-line over historical data, with an extremely efficient generative model, updated on-line according to the user behavior. Qualitative and quantitative analyses show promising results.