ImageNet

ImageNet is an image dataset organized according to the WordNet hierarchy. Each meaningful concept in WordNet, possibly described by multiple words or word phrases, is called a "synonym set" or "synset". There are more than 100,000 synsets in WordNet, majority of them are nouns (80,000+). In ImageNet, we aim to provide on average 1000 images to illustrate each synset. Images of each concept are quality-controlled and human-annotated.

3D Object Recognition  PPT

We propose a novel probabilistic framework for learning visual models of 3D object categories by combining appearance information and geometric constraints. Objects are represented as a coherent ensemble of parts that are consistent under 3D viewpoint transformations. Each part is a collection of salient image features. A generative framework is used for learning a model that captures the relative position of parts within each of the discretized viewpoints.

Sky Recognition

Reliable sky detection is a difficult task due to its large variance in color, texture, illumination and location. We propose a method combining global and local evidence for reliable sky detection. At the global scale, a shiftable kernel is used to incorporate spatial information which is robust to image translation; at the local scale, a boosting classifier is train to do pixel-wise segmentation. Finally, a fusion classifier is proposed to combine the output of two classifiers so that they correct each other.

Automatic Chinese Couplet Generation

Chinese Couplet is a Chinese article style, which is composed of a pair of sentences with the same number of characters matching both in sound and senses. The Automatic Chinese Couplet Generation system is able to automatically propose a list of matching second sentences if the user would give a first one. This system integrates Data Mining, Information Retrieval and several Natural Language Processing techiques.