|
|
--- |
|
|
task_categories: |
|
|
- audio-to-audio |
|
|
size_categories: |
|
|
- 100B<n<1T |
|
|
--- |
|
|
# SFR-v1: FM Reception Speech Dataset |
|
|
|
|
|
## Dataset Summary |
|
|
|
|
|
**SFR-v1** consists of segmented 10-second FM reception audio clips accompanied by frame-level RSSI (Received Signal Strength Indicator) traces. This dataset is designed to support research in hybrid digital-analog communication, channel-aware audio recovery, speech enhancement, and FM channel modeling. It includes multiple content types (news, music, speech, male and female voices) under varying channel quality conditions. |
|
|
|
|
|
## Audio Parameters |
|
|
|
|
|
* **Sample rate**: 24 kHz |
|
|
* **Bit depth**: 16-bit PCM |
|
|
* **Channels**: Mono |
|
|
|
|
|
## RSSI Parameters |
|
|
|
|
|
* **Acquisition**: a probe signal used solely for RSSI acquisition and not part of the audio |
|
|
* **Original interval**: \~60 ms per measurement (with timestamp) |
|
|
* **Resampling**: interpolated to 50 ms per step, yielding 200 RSSI values per 10-second clip |
|
|
* **Raw range**: approximately –65 dBm to –105 dBm |
|
|
* **Normalization**: linearly mapped to \[4, 20], where higher values indicate poorer channel quality |
|
|
* **Alignment**: RSSI sequences are frame-aligned with each audio segment |
|
|
|
|
|
## License |
|
|
|
|
|
Released under CC-BY-NC 4.0. |
|
|
|
|
|
## Citation |
|
|
|
|
|
Citation will be provided upon publication. |
