Frequently Asked Questions
Last Revised: August 26, 2025
Table of Contents
Paper Submission Guidelines
Q: What is the page limit likely to be for paper submissions?
A: Workshop papers should follow the ICASSP 2026 paper style, format, and length guidelines. Detailed guidelines can be found on the ICASSP 2026 paper kit website: https://cmsworkshops.com/ICASSP2026/papers/paper_kit.php.
Q: Are there any particular compute, latency, or bitrate constraints (like for the 2025 LRAC Challenge papers)?
A: In brief, there are no specific compute, latency, or bitrate constraints for non-Challenge paper submissions to the Workshop. However, we do ask that submissions maintain a clear connection to the workshop’s central theme of low-resource, efficient codecs.
The Workshop’s broader focus is on ensuring that low-resource neural speech codecs are lightweight and suitable for deployment in real-world products and applications. The precise requirements can vary widely depending on the use case. For instance, speech codecs for real-time video conferencing often require strict end-to-end latency in the range of 10–50 ms, whereas codecs used as feature extractors for AI agents running on cloud GPUs may tolerate overall latencies up to 500 ms.
As a point of distinction between the LRAC Challenge and the LRAC Workshop: the Challenge has its own defined constraints and evaluation criteria. On the other hand, while the Workshop welcomes Challenge papers, it has a broader scope and encourages submissions that explore a variety of application scenarios, system requirements, and topics, provided the primary focus remains on low-resource, efficient neural codecs.
If you have further questions, please reach out to us: lrac-workshop@cisco.com