Argus (Aegis) —
Conv2D Iris Encoder (Production)
Production deployment of the Argus iris recognition encoder on Hailo-10H. This is the high-throughput variant running on the NexusWatch perception platform, achieving 3,563 FPS — sufficient for continuous multi-camera iris scanning in high-security access control deployments.
Abstract
Background
Production deployment of the Argus iris recognition encoder on Hailo-10H. This is the high-throughput variant running on the NexusWatch perception platform, achieving 3,563 FPS — sufficient for continuous multi-camera iris scanning in high-security access control deployments.
Approach
The model was trained in AxonML (a pure-Rust deep learning framework) and compiled through the Hailo Dataflow Compiler (DFC 5.3.0) targeting Hailo-10H silicon. Post-training INT8 quantization was applied during the DFC compilation pass with production telemetry calibration data. The resulting Hailo Executable Format (HEF) binary executes on Hailo’s fixed-function dataflow architecture with deterministic latency and zero framework overhead at the edge.
Results
On production hardware (Hailo-10H (NexusWatch, FW 5.3.0)), Argus (Aegis) achieves 3,563 FPS with 45.7 °C average die temperature (max 46.6 °C).
Conclusion
Argus (Aegis) is production-ready as a single HEF binary deployed to edge devices with no external dependencies beyond the HailoRT vendor runtime. The model meets real-time latency requirements for its target biometric security application.
| Model | Argus (Aegis) |
| Domain | Biometric Security |
| Architecture | Conv2D Iris Encoder (Production) |
| Target silicon | Hailo-10H |
| Measured on | Hailo-10H (NexusWatch, FW 5.3.0) |
| DFC compiler | 5.3.0 |
| Framework | AxonML v0.6 (pure-Rust, CUDA + CPU backends) |
| Author | Andrew Jewell Sr. · ORCID 0009-0005-2158-7060 |
| Organization | AutomataNexus LLC · Fort Wayne, Indiana |
Executive overview
Production deployment of the Argus iris recognition encoder on Hailo-10H. This is the high-throughput variant running on the NexusWatch perception platform, achieving 3,563 FPS — sufficient for continuous multi-camera iris scanning in high-security access control deployments.
Network I/O
Input: NIR iris image [1, 1, 64, 64]. Output: 512-dim identity embedding.
Architecture
Conv2D Iris Encoder (Production)
7-block progressive Conv2D encoder identical to the H8 variant. On Hailo-10H, the larger dataflow engine (40 TOPS vs 26 TOPS) provides higher throughput but the model is memory-bound rather than compute-bound, so the speedup is moderate (3,563 vs 4,804 FPS on H8 — H8 is actually faster here due to better resource utilization on the smaller model).
Compilation constraints
All AxonML models targeting Hailo silicon are compiled under the fixed-function dataflow constraints: no dynamic control flow, no variable-length dimensions, all activations representable in INT8 after calibration, and no operations requiring dedicated softmax hardware (replaced with ReLU gating or depthwise convolution equivalents where necessary).
Silicon performance
Measured on production hardware via hailortcli benchmark with 5-second sustained inference. Device: Hailo-10H (NexusWatch, FW 5.3.0).
| Metric | Measured Value |
|---|---|
| FPS (streaming) | 3,562.69 |
| Die Temperature (mean) | 45.67 °C |
| Die Temperature (min) | 43.13 °C |
| Die Temperature (max) | 46.57 °C |
| Quantization | INT8 (post-training, DFC calibration) |
| DFC Compiler | 5.3.0 |
| HailoRT | 5.3.0 |
| Measured On | Hailo-10H (NexusWatch, FW 5.3.0) |
Deployment
Deployed as a single HEF binary. No ONNX runtime, TensorFlow Lite, or Python inference stack required at the edge.
| Target silicon | Hailo-10H |
| Measured on | Hailo-10H (NexusWatch, FW 5.3.0) |
| DFC compiler | 5.3.0 |
| Quantization | INT8 (post-training, production telemetry calibration) |
| Runtime | HailoRT (vendor runtime) |
| Edge platform | Raspberry Pi 5 + Hailo AI HAT+ (M.2 Key M) |
Deployment procedure
Copy the .hef binary to the target device. hailortcli run loads the HEF directly into the Hailo-10H dataflow engine over PCIe. Inference begins immediately with deterministic per-frame latency. No model conversion, graph optimization, or warmup phase required.
References
- Jewell, A. (2026). AxonML: A Pure-Rust Deep Learning Framework for Edge Inference. AutomataNexus LLC. Technical whitepaper.
- Hailo Technologies Ltd. (2024). Hailo Dataflow Compiler User Guide. DFC v5.3.0.
- Hailo Technologies Ltd. (2024). Hailo-10H Product Datasheet.
AutomataNexus LLC · Fort Wayne, Indiana · andrew.jewellsr@automatanexus.com
Andrew Jewell Sr. · ORCID 0009-0005-2158-7060
May 2026 · All rights reserved.