Compound facial expressions of emotion

By Shichuan Du, Yong Tao, Aleix M. Martinez. Proceedings of the National Academy of Sciences of the United States of America, 2014

803 citations recorded by the source corpus.

Subfields: Cognitive Neuroscience, Sensory Systems, Social Psychology

View Paper: https://www.semanticscholar.org/paper/45deb9d034e5c99de0e36e4c2a9a4f4277d1497f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Sensory Systems

In Social Psychology

Citations in the corpus

Survey on Emotional Body Gesture Recognition (2018, methodology): “While this model is rarely used in affective computing compound emotions have gained relative visibility [53].”
A Brief Review of Facial Emotion Recognition Based on Visual Information (2018, background): “[17] introduced 22 emotions, including seven basic emotions, 12 compound emotions most typically expressed by humans, and three additional emotions (appall, hate, and awe).”
Describing Common Human Visual Actions in Images (2015, methodology): “2Despite the disagreement on the fact that humans might have basic discrete emotions [5, 19], we adopt Ekman’s 6 basic emotions [6] for this study as we are interested in a high level description of subject’s emotional state.”
Reliable Crowdsourcing and Deep Locality-Preserving Learning for Expression Recognition in the Wild (2017, background): “2 million labels of 29672 great-diverse facial images downloaded from the Internet, these Real-world Affective Faces (RAF)1 are naturally categorized into two types: basic expression with single-modal distribution and compound emotions with bimodal distribution, an observation supporting a recent gr”
paGAN (2018, neutral): “• 230 subjects with 22 expressions each from the compound facial expressions (CFE) dataset [Du et al. 2014].”
Facial Expression Recognition with Inconsistently Annotated Datasets (2018, background): “Experiment settings To evaluate the methods’ generalization ability on data under the unseen condition, cross-dataset evaluation protocol was applied for SFEW, CK+, CFEE, MMI, and Oulu-CASIA datasets.”
How We've Taught Algorithms to See Identity: Constructing Race and Gender in Image Databases for Facial Analysis (2020, methodology): “—CFEE [39] In this instance, “female” is used as a default gender, implying there must be another gender accounted for in the demographic distribution (most likely “male,” given binary trends).”
GANimation: Anatomically-aware Facial Animation from a Single Image (2018, background): “For example, the facial expression for fear is generally produced with activations: Inner Brow Raiser (AU1), Outer Brow Raiser (AU2), Brow Lowerer (AU4), Upper Lid Raiser (AU5), Lid Tightener (AU7), Lip Stretcher (AU20) and Jaw Drop (AU26) [5].”
AffectNet: A Database for Facial Expression, Valence, and Arousal Computing in the Wild (2017, methodology): “Images were labeled as one of the 23 (basic or compound) emotion categories defined in [6] based on AUs.”
EMOCA: Emotion Driven Monocular Face Capture and Animation (2022, neutral): “) or compound expression categories [21] (e.”
Choosing prediction over explanation in psychology: Lessons from machine learning (2017, background): “For example, Du, Tao, and Martinez (2014) proposed that, in addition to the six “basic” facial emotional expressions of happiness, surprise, sadness, anger, fear, and disgust (Ekman, 1992), human beings can reliably produce and detect additional compound expressions.”
Knowledge Augmented Deep Neural Networks for Joint Facial Expression and Action Unit Recognition (2020, background): “Besides, Du et al [6] quantitatively analyzed the relationships among expression and AUs based on their studies on different subjects and reported the probabilities for variant AUs under each expression.”
Deep Region and Multi-label Learning for Facial Action Unit Detection (2016, methodology): “Region learning (RL): Conventional methods for AU detection utilize geometric features [4, 19], appearance features [2,6,13] or both [3].”
EmotioNet: An Accurate, Real-Time Algorithm for the Automatic Annotation of a Million Facial Expressions in the Wild (2016, methodology): “Specifically, we used three of the above-defined databases – CFEE, DISFA and CK+ – and run a leave-onedatabase out test.”
Automated Facial Trait Judgment and Election Outcome Prediction: Social Dimensions of Face (2015, background): “[31], emotional state [11, 10], and identity [38, 22], each of which has been studied intensively both in human perception and computer vision.”
Expression, Affect, Action Unit Recognition: Aff-Wild2, Multi-Task Learning and ArcFace (2019, neutral): “We note that the Expr Set of images can be extended to also contain AU annotations, according to Table 1 of [10].”
Multimodal Spontaneous Emotion Corpus for Human Behavior Analysis (2016, background): “According to the compound emotions theory [7], the surprised feeling could be positive or negative.”
Deep Facial Expression Recognition: A Survey (2018, background): “[51], who argued that several facial emotions are actually combinations ofmore than one basic expression.”
Reliable Crowdsourcing and Deep Locality-Preserving Learning for Unconstrained Facial Expression Recognition (2019, result): “This observation supports a recent finding in the lab-controlled condition [14].”