emo-affectnet-jetson-orin

EMO-AffectNet (VGGFace2 ResNet50) — Jetson Orin Nano FP16

7-class facial emotion classifier (Neutral / Happy / Sad / Surprise / Fear / Disgust / Anger). Keras .h5 + ONNX + TensorRT FP16 engine with dynamic batch (1–16).

This is a private mirror prepared for the robot_perception_scoring project (github.com/JSBAICenter/Nvidia_Edge_work_arounds, branch orin_yolo). It bundles the original source weights together with an ONNX export and a TensorRT FP16 engine built on a Jetson Orin Nano (8 GB).

Files in this repo

How to load

• keras_h5: tf.keras.models.load_model('weights_0_66_37_wo_gl.h5')
• onnx: onnxruntime.InferenceSession (any CUDA-capable host)
• engine: tensorrt.Runtime.deserialize_cuda_engine (Jetson Orin Nano only)

The .engine artifact is hardware-specific to Jetson Orin Nano (Ampere sm_87, TensorRT 10.3, CUDA 12.6, FP16). It will not load on other Jetson families, other TensorRT versions, or x86 GPUs. Other users should rebuild from the .onnx via python -m tools.build_engines in the linked GitHub repo.

Build environment

• JetPack tegra release: # R36 (release), REVISION: 4.7, GCID: 42132812, BOARD: generic, EABI: aarch64, DATE: Thu Sep 18 22:54:44 UTC 2025
• CUDA: Cuda compilation tools, release 12.6, V12.6.68
• TensorRT: 10.3.0

Attribution

EMO-AffectNet trained over the VGGFace2 ResNet50 backbone. Preprocessing matches keras-vggface preprocess_input(version=2): 224x224 BGR, per-channel mean subtraction (B-91.5, G-103.9, R-131.1).

License

Per the original EMO-AffectNet authors. Re-distribution beyond this private mirror should be cleared with them.