Information extraction from image sequences of real-world facial expressions

Gu, Haisong; Ji, Qiang

doi:10.1007/s00138-004-0161-6

Cited by 22 publications

(18 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However detecting these facial expressions in the less constrained environments of real applications is a much more challenging problem which is just beginning to be explored. There have also been a few tentative efforts to detect cognitive and psychological states like interest (El Kaliouby & Robinson, 2004), pain , and fatigue (Gu & Ji, 2005). In sign judgment approaches (Cohn & Ekman, 2005), a widely used method for manual labeling of facial actions is the Facial Action Coding System (FACS; Ekman & Friesen, 1978, Ekman et al, 2002.…”

Section: Level Of Description: Action Units and Emotionsmentioning

confidence: 99%

“…To address the limitations inherent in optical flow techniques such as the accumulation of error and the sensitivity to noise, occlusion, clutter, and changes in illumination, several researchers used sequential state estimation techniques to track facial feature points in image sequences. Both, Zhang and Ji (2005) and Gu and Ji (2005) used facial point tracking based on a Kalman filtering scheme, which is the traditional tool for solving sequential state problems. The derivation of the Kalman filter is based on a state-space model (Kalman, 1960), governed by two assumptions: (i) linearity of the model and (ii) Gaussianity of both the dynamic noise in the process equation and the measurement noise in the measurement equation.…”

Section: Facial Point Trackingmentioning

confidence: 99%

“…Thus, if the face undergoes a sudden or rapid movement, the prediction of features positions from Kalman filtering will be significantly off. To handle this problem, Zhang and Ji (2005) and Gu and Ji (2005) used the information about the IR-camera detected pupil location together with the output of Kalman filtering to predict facial features positions in the next frame of an input face video. To overcome these limitations of the classical Kalman filter and its extended form in general, particle filters wereproposed.…”

Section: Facial Point Trackingmentioning

confidence: 99%

See 2 more Smart Citations

Machine Analysis of Facial Expressions

Pantić¹,

Bartlett²

2007

Face Recognition

176

156

View full text Add to dashboard Cite

Section: Level Of Description: Action Units and Emotionsmentioning

confidence: 99%

Section: Facial Point Trackingmentioning

confidence: 99%

Section: Facial Point Trackingmentioning

confidence: 99%

See 1 more Smart Citation

Machine Analysis of Facial Expressions

Pantić¹,

Bartlett²

2007

Face Recognition

176

156

View full text Add to dashboard Cite

“…Similarly, the most advanced approaches to facial feature tracking are based on Kalman (e.g. [23]) and particle filtering tracking schemes (e.g. [61]).…”

Section: State Of the Fieldmentioning

confidence: 99%

Human Computing and Machine Understanding of Human Behavior: A Survey

Pantić

Pentland

Nijholt

et al.

Artifical Intelligence for Human Computing

285

214

View full text Add to dashboard Cite

A widely accepted prediction is that computing will move to the background, weaving itself into the fabric of our everyday living spaces and projecting the human user into the foreground. If this prediction is to come true, then next generation computing, which we will call human computing, should be about anticipatory user interfaces that should be human-centered, built for humans based on human models. They should transcend the traditional keyboard and mouse to include natural, human-like interactive functions including understanding and emulating certain human behaviors such as affective and social signaling. This article discusses a number of components of human behavior, how they might be integrated into computers, and how far we are from realizing the front end of human computing, that is, how far are we from enabling computers to understand human behavior. Categories and Subject Descriptors General TermsAlgorithms, Theory, Performance KeywordsMultimodal user interfaces, Anticipatory user interfaces, Sociallyaware computing, Affective computing HUMAN COMPUTINGFuturistic movies often contain visions of human environments of the future. Fitted out with arrays of intelligent, yet invisible devices, homes, transportation means and working spaces of the future can anticipate every need of their inhabitants (Fig. 1). This vision of the future is often referred to as "ubiquitous computing" [66] or "ambient intelligence" [1]. In this vision of the future, humans will be surrounded by intelligent interfaces that are supported by computing and networking technology embedded in all kinds of objects in the environment and that are sensitive and responsive to the presence of different individuals in seamless and unobtrusive way. This assumes a shift in computing -from desktop computers to a multiplicity of smart computing devices diffused into our environment. It assumes that computing will move to the background, weave itself into the fabric of everyday living spaces and disappear from the foreground, projecting the human user into it. However, as computing devices disappear from the scene, become invisible, weaved into our environment, a new set of issues is created concerning the interaction between this technology and humans [35] One way of tackling these problems is to move away from computer-centered designs toward human-centered designs for human computer interaction (HCI). The former involve usually the conventional interface devices like keyboard, mouse, and visual displays, and assume that the human will be explicit, unambiguous and fully attentive while controlling information and command flow. This kind of interfacing and categorical computing works well for context-independent tasks like making plane reservations and buying and selling stocks. However, it is utterly inappropriate for interacting with each of the (possibly hundreds) computer systems diffused throughout future smart environments and aimed at improving the quality of life by anticipating the users needs. The key to human computing an...

show abstract

“…Even though the automatic recognition of the six basic emotions from face images and image sequences is considered largely solved, reports on novel approaches are published even to date (e.g., [13]- [16]). Exceptions from this overall state of the art in machine analysis of human facial affect include few tentative efforts to detect cognitive and psychological states like interest [17], pain [18], [19], and fatigue [20]. Facial emotion recognition from 2D images is well studied field but lack of real-time method that estimates features even low quality images.…”

Section: Introductionmentioning

confidence: 99%