[Paper Review] A transformer-based approach to video frame-level prediction in Affective Behaviour Analysis In-the-wild
The paper presents a transformer-based model for frame-level emotion classification in Affective Behavior Analysis In-the-wild, achieving 0.4775 on Aff-Wild2 validation.
In recent years, transformer architecture has been a dominating paradigm in many applications, including affective computing. In this report, we propose our transformer-based model to handle Emotion Classification Task in the 5th Affective Behavior Analysis In-the-wild Competition. By leveraging the attentive model and the synthetic dataset, we attain a score of 0.4775 on the validation set of Aff-Wild2, the dataset provided by the organizer.
Motivation & Objective
- Motivate the use of transformer architectures for frame-level affective computing in-the-wild settings.
- Leverage a synthetic dataset to train a frame-level emotion classifier.
- Evaluate the model on the Aff-Wild2 validation set provided by the competition organizer.
Proposed method
- Proposes a transformer-based model for frame-level prediction in affective behavior analysis in-the-wild.
- Utilizes attentive mechanisms to process video frame sequences.
- Trains on a synthetic dataset and evaluates on Aff-Wild2 validation data.
- Aims to improve Emotion Classification Task performance in the ABIA 5th competition.
Experimental results
Research questions
- RQ1Can a transformer-based architecture effectively perform frame-level emotion classification in-the-wild video data?
- RQ2What impact does using a synthetic dataset have on performance for the ABIA emotion classification task?
- RQ3What validation performance can be achieved on Aff-Wild2 with the proposed framework?
Key findings
- Achieves a validation score of 0.4775 on Aff-Wild2.
- Demonstrates the effectiveness of attentive transformer models for frame-level affective prediction.
- Shows competitive results within the ABIA 5th competition setup using synthetic training data.
Better researchstarts right now
From reading papers to final review, dramatically reduce your research time.
No credit card · Free plan available
This review was created by AI and reviewed by human editors.