Advances in Emotional Picture Classification

Advances in Emotional Picture Classification

Yu-Jin Zhang (Department of Electronic Engineering, Tsinghua University, China)
Copyright: © 2015 |Pages: 11
DOI: 10.4018/978-1-4666-5888-2.ch048

Chapter Preview



To bridge the “semantic gap” between low-level features and high-level emotional concepts, a middle-level is inserted. Unlike the traditional machine learning methods that directly establish the mapping relationship from the low-level features to high-level emotional concepts, a new concept called latent emotional semantic factors is introduced. The unsupervised probabilistic Latent Semantic Analysis (pLSA) model is used to discover the latent emotional semantic factors.

The procedure is depicted in Figure 1. It starts from “feature extraction,” via “latent emotional factors,” to “emotion classification.” They are corresponding to “low picture feature,” “middle semantic level,” and “high emotion concept,” respectively.

Figure 1.

Three levels of emotional picture classification

Key Terms in this Chapter

Content-Based Visual Information Retrieval (CBVIR): A combination of CBIR and CBVR.

Image Classification: Aims at associating different images with some semantic labels to represent the image contents abstractly. To achieve this goal, various machine learning and pattern recognition techniques could be used.

Content-Based Video Retrieval (CBVR): A process framework for efficiently retrieving required clip from video. The retrieval relies on the organization of video and nonlinear search techniques.

Facial Expression: A form of nonverbal communication. A facial expression is one or more motions or positions of the muscles beneath the skin of the face. These movements convey the emotional state of an individual to observers. Facial recognition is often an emotional experience for the brain.

Emotion: A positive or negative experience that is associated with a particular pattern of physiological activity. In psychology and philosophy, a subjective experience is characterized by psychophysiological expressions, biological reactions, and mental states. It is often the driving force behind motivation.

Scale-Invariant Feature Transform (SIFT): An algorithm in to detect and describe local features in images, and sometimes, the local feature itself. SIFT can be seen as a method for image feature generation transforms an image into a large collection of feature vectors, each of which is invariant to image translation, scaling, and rotation, partially invariant to illumination changes and robust to local geometric distortion. It has found application in various area of image processing, image analysis and image understanding, such as object recognition, stereo matching, video tracking, etc.

Content-Based Image Retrieval (CBIR): A process framework for efficiently retrieving images from a collection by similarity. The retrieval relies on extracting the appropriate characteristic quantities describing the desired contents of images. In addition, suitable querying, matching, indexing and searching techniques are required.

Complete Chapter List

Search this Book: