Reproduce “Implementation of Zero-shot Semantic Communication on Software Defined Radio” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Thanh Le, Arif Dataesatu, Homare Murakami, Takeshi Matsumura
The system demonstrates zero-shot classification over a radio link by sending vision-language embeddings for receiver-side comparison with text prompts, allowing labels to change without retraining. On 100 CIFAR-10 images, CLIP reaches 82% top-1 accuracy at 22.3 dB, while JPEG + 16-QAM scores 0%. Semantic transmission uses 9× fewer channel uses per image; radios were coax-connected, not over the air.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.