metadata

license: bsd-3-clause
metrics:
  - accuracy
base_model:
  - MIT/ast-finetuned-audioset-10-10-0.4593
pipeline_tag: audio-classification
tags:
  - signal
  - radio
  - rf
  - emission
  - spectrum

Audio Spectrogram Transformer finetuned on SIGID wiki for Radio Signal Classification

This model is a fine-tuned version of MIT/ast-finetuned-audioset-10-10-0.4593 on radio signal recordings documented in SIGID wiki. It was built by ATDI's AI team to provide a base model for radio signal classification.

It achieves the following results on the evaluation set:

Accuracy: 0.99
Validation Loss: 0.05

Dataset

We built our own dataset using SIGID wiki. The model is trained to recognize the following radio systems:

4G LTE Network
5G "New Radio" cellular network - Downlink
Aircraft Communications Addressing and Reporting System (ACARS)
Amplitude Modulation (AM)
Automatic Identification System (AIS)
Automatic Link Set-up (ALIS)
Automatic Picture Transmission (APT)
Bluetooth
Differential Global Positioning System (DGPS)
Digital Audio Broadcasting Plus (DAB+)
Digital Mobile Radio (DMR)
Digital Video Broadcasting — Terrestrial (DVB-T)
High Frequency Data Link (HFDL)
Instrument Landing System
M20 Radiosonde
Morse Code (CW)
Non-Directional Beacon (NDB)
Radar altimeter
STANAG 5065
Secondary surveillance radar (SSR)
Single Sideband Voice
Tetrapol
VHF Data Link - Mode 2 (VDL-M2)
VHF Omnidirectional Range (VOR)

Usage

You can use the raw model for classifying signals into one of the SIGID wiki classes specified above and in config.json. You can also fine-tune this model on your own radio signal dataset to make it more specific.