---
license: apache-2.0
library_name: compressionkit
pipeline_tag: other
tags:
  - compressionkit
  - signal-compression
  - ppg
  - spiht
  - wavelet
  - edge-ai
  - dsp
---

# compressionkit-ppg-spiht-32x

A **PPG** signal compression codec built on wavelet + SPIHT + arithmetic coding. **DSP-only — no trained weights.** The deployable artifact is the bitstream contract plus a portable C99 reference implementation.

## Operating point

| Field | Value |
|-------|-------|
| Modality | PPG |
| Sample rate | 64 Hz |
| Frame size | 320 samples |
| Target CR | 32x |
| Wavelet | `coif5` |
| DWT levels | 6 |
| Bit budget | 160 bits/frame |
| Entropy coder | arithmetic coding |

## Quality metrics

### Fidelity & Robustness

Both fidelity yardsticks are reported so the codec is judged fairly: **faithfulness** is PRD vs the recorded (still-noisy) input, while **truth fidelity** is PRD vs clean ground truth. Lower is better.

| Metric | Value |
|--------|-------|
| Truth PRD vs clean (%) | 19.35 |
| Truth PRD at native noise (%) | 40.07 |
| Faithful PRD vs input (%) | 20.38 |
| PRD degradation slope (PRD%/dB) | 4.18 |
| PRD at 0 dB SNR (%) | 78.25 |
| PRD at -6 dB SNR (%) | 106.31 |
| Pure-noise imprint autocorr | 0.2850 |

### Time Domain

_PRD here is faithfulness (vs the recorded input); see **Fidelity & Robustness** above for the clean-truth and noise-regime view._

| Metric | Mean | Median | P90 |
|--------|------|--------|-----|
| PRD vs input — faithfulness (%) | 20.3765 | 19.8541 | 26.5802 |
| RMSE | 0.2038 | 0.1985 | 0.2658 |
| Cosine Similarity | 0.9783 | 0.9804 | 0.9889 |

### Spectral

- **Band Total Relative Error (median):** 0.1120

### Bitrate


## Python quickstart

```python
from compressionkit.runtime import load_codec

codec = load_codec("Ambiq/compressionkit-ppg-spiht-32x")
enc = codec.compress(frame)   # frame: (frame_size,) float32
recon = codec.decompress(enc)
```

## C quickstart

```c
#include "spiht_app_config.h"

float frame[APP_SPIHT_FRAME_SIZE];
uint8_t bitstream[APP_SPIHT_MAX_BYTES];
/* ... fill frame from sensor (post-denoise, if hybrid) ... */
size_t nbits = spiht_encode_frame(&enc, bitstream, APP_SPIHT_MAX_BITS);
```

## Files

| File | Description |
|------|-------------|
| `config.json` | Deploy manifest (`family: "spiht"`) |
| `spiht_config.json` | Codec parameters (language-neutral) |
| `sample_stimulus.npz` | Synthetic test frames |
| `reference_vectors.npz` | Reference encode/decode vectors |
| `c_sources/spiht.[ch]` | Portable C99 reference |
| `c_sources/spiht_app_config.h` | Codec-specific defines |
| `model_card.json` | Provenance metadata |
| `scorecard.json` | Frozen evaluation summary |

## Dataset & license

Evaluation data provenance is not recorded in this package; sample data uses synthetic physiokit waveforms only — no patient data is redistributed.

Codec source released under the **APACHE-2.0** license.

## Citation

```bibtex
@software{compressionkit,
  author = {Ambiq AI},
  title = {compressionKIT: Signal Compression for Edge AI},
  url = {https://github.com/AmbiqAI/compressionkit}
}
```