MétaCan
Menu
Back to cohort
Record W2962980504 · doi:10.2196/13722

Opportunities and Pitfalls in Applying Emotion Recognition Software for Persons With a Visual Impairment: Simulated Real Life Conversations

2019· article· en· W2962980504 on OpenAlexvenueno aff
Hendrik Buimer, Renske Schellens, Tjerk Kostelijk, Abdellatif Nemri, Yan Zhao, Thea van der Geest, Richard van Wezel

Bibliographic record

VenueJMIR mhealth and uhealth · 2019
Typearticle
Languageen
FieldNeuroscience
TopicTactile and Sensory Interactions
Canadian institutionsnot available
Fundersnot available
KeywordsPsychologyFacial expressionGestureNonverbal communicationEmotion recognitionDisgustSession (web analytics)Cognitive psychologyCommunicationComputer scienceSocial psychologyArtificial intelligenceAnger

Abstract

fetched live from OpenAlex

BACKGROUND: A large part of the communication cues exchanged between persons is nonverbal. Persons with a visual impairment are often unable to perceive these cues, such as gestures or facial expression of emotions. In a previous study, we have determined that visually impaired persons can increase their ability to recognize facial expressions of emotions from validated pictures and videos by using an emotion recognition system that signals vibrotactile cues associated with one of the six basic emotions. OBJECTIVE: The aim of this study was to determine whether the previously tested emotion recognition system worked equally well in realistic situations and under controlled laboratory conditions. METHODS: The emotion recognition system consists of a camera mounted on spectacles, a tablet running facial emotion recognition software, and a waist belt with vibrotactile stimulators to provide haptic feedback representing Ekman's six universal emotions. A total of 8 visually impaired persons (4 females and 4 males; mean age 46.75 years, age range 28-66 years) participated in two training sessions followed by one experimental session. During the experiment, participants engaged in two 15 minute conversations, in one of which they wore the emotion recognition system. To conclude the study, exit interviews were conducted to assess the experiences of the participants. Due to technical issues with the registration of the emotion recognition software, only 6 participants were included in the video analysis. RESULTS: We found that participants were quickly able to learn, distinguish, and remember vibrotactile signals associated with the six emotions. A total of 4 participants felt that they were able to use the vibrotactile signals in the conversation. Moreover, 5 out of the 6 participants had no difficulties in keeping the camera focused on the conversation partner. The emotion recognition was very accurate in detecting happiness but performed unsatisfactorily in recognizing the other five universal emotions. CONCLUSIONS: The system requires some essential improvements in performance and wearability before it is ready to support visually impaired persons in their daily life interactions. Nevertheless, the participants saw potential in the system as an assistive technology, assuming their user requirements can be met.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.002
metaresearch head score (Gemma)0.010
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Observational · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.002
Threshold uncertainty score0.013

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0020.010
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0000.000
Science and technology studies0.0010.000
Scholarly communication0.0010.001
Open science0.0010.001
Research integrity0.0010.000
Insufficient payload (model declined to judge)0.0020.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.117
GPT teacher head0.368
Teacher spread0.251 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designObservational
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations4
Published2019
Admission routes1
Has abstractyes

Explore more

Same venueJMIR mhealth and uhealthSame topicTactile and Sensory InteractionsFrench-language works237,207