Abstract
Real-time media moderation on resource-constrained devices, for example, smart televisions, may present computational challenges, while cloud-based solutions can introduce latency and privacy concerns. A system can distribute computational workloads for content moderation across a local area network, such as a Wi-Fi network. A primary media device can operate a lightweight trigger agent to heuristically scan for potential anomalies in a media stream. Upon detection, a segment of the media may be offloaded to one or more available secondary devices, for instance, smartphones or personal computers, on the network. These secondary devices can perform computationally intensive analysis using advanced models and return metadata to the primary device. The primary device can use a consensus protocol to determine if a modification, such as blurring or muting, could be applied. This approach may provide low-latency, privacy-preserving moderation by processing content locally and can help mitigate performance degradation on the primary playback device.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Kushwaha, Vivek, "Asymmetric Workload Offloading for Real-Time Media Moderation on a Local Device Network", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/12085