Tech Papers 2021: This paper presents a lightweight CNN for subtle facial expression analysis in images and videos, enabling applications like expression-based search and actor summaries.
Abstract
Existing computer vision based emotion recognition systems are trained to classify images of faces into a very limited number of emotions. In this work, we take a different approach and train a convolutional neural network that is able to distinguish subtle differences in facial expressions appearing in both individual images and videos.
For this effect, we learn a feature embedding network which maps a facial image into a position in embedding space, such that it is positioned close to similar facial expressions compared with other expressions. We train the feature embedding using triplet loss on the publicly available FEC dataset. The proposed facial expression model is lightweight (4.7M parameters), and obtains a triplet prediction accuracy of 84.5% -- very close to the average human performance of 86.2%.
Exclusive Content
This article is available with a Technical Paper Pass
Opportunities for emerging 5G and wifi 6E technology in modern wireless production
This paper examines the changing regulatory framework and the complex technical choices now available to broadcasters for modern wireless IP production.
Leveraging AI to reduce technical expertise in media production and optimise workflows
Tech Papers 2025: This paper presents a series of PoCs that leverage AI to streamline broadcasting gallery operations, facilitate remote collaboration and enhance media production workflows.
Automatic quality control of broadcast audio
Tech Papers 2025: This paper describes work undertaken as part of the AQUA project funded by InnovateUK to address shortfalls in automated audio QC processes with an automated software solution for both production and distribution of audio content on premises or in the cloud.
Demonstration of AI-based fancam production for the Kohaku Uta Gassen using 8K cameras and VVERTIGO post-production pipeline
Tech Papers 2025: This paper details a successful demonstration of an AI-based fancam production pipeline that uses 8K cameras and the VVERTIGO post-production system to automatically generate personalized video content for the Kohaku Uta Gassen.
EBU Neo - a sophisticated multilingual chatbot for a trusted news ecosystem exploration
Tech Papers 2025: The paper introduces NEO, a sophisticated multilingual chatbot designed to support a trusted news ecosystem.
