AIPV1EMI0001 — YOLOv7-tiny Parking Detector (EdgeAI Station)

Model weights and build artifacts for the AIPV1EMI0001 AI Processor (YOLOv7-tiny parking-lot detector: car, person, truck — 3 classes, 640×640 input).

Source code, build scripts and the step-by-step guide live in the GitHub repo: edge-ai/AIPV1EMI0001

Files

File	Description
yolov7tiny-parking-rev3-640.onnx	Primary artifact. ONNX export with EfficientNMS_TRT node — input for the TensorRT engine build (bin/build.sh). Downloaded automatically by the notebook.
yolov7-fp16-1x8x8-jp6-l4t36.3-arm64.engine	Prebuilt TensorRT FP16 engine — JetPack 6 / L4T R36.3 / aarch64 only. Optional shortcut; rebuild from ONNX on any other platform.
SHA256SUMS	Checksums for integrity verification after download

Usage

The localdev.ipynb notebook in the GitHub repo downloads the ONNX automatically and verifies it against SHA256SUMS before building the TensorRT engine.

Attribution & License

The model architecture is YOLOv7 by WongKinYiu et al.:

Chien-Yao Wang, Alexey Bochkovskiy, Hong-Yuan Mark Liao. "YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors." CVPR 2023. https://github.com/WongKinYiu/yolov7 — MIT License.

The weights published here were custom-trained and reparameterized by EdgeMatrix (parking-lot dataset, 3 classes) on top of the YOLOv7-tiny architecture, and are released under Apache-2.0. The upstream YOLOv7 code remains under its MIT license.