rock_paper_scissors_cnn: Verified on the ECS-DoT EVB

Two-convolution strided CNN that classifies 32x32 greyscale images of a hand showing rock, paper or scissors. A 9x9 stride-2 first convolution brings the map to 12x12, a 3x3 stride-2 convolution to 6x6, then one Linear classifier. Zero-mean weights, BatchNorm folded, ReLU only. Validated on the ECS-DoT EVB.

Compiled with ECS-Compiler in ECS-Studio 1.140.0 and verified on a physical ECS-DoT EVB.

Getting Started

This repository is a ready-to-use model folder for ECS-Compiler, the AI pipeline of ECS-Studio. It holds the ONNX model and its calibration and evaluation data. The verified result was built with the default ECS-Compiler settings; keep them unchanged to reproduce it.

Compile the model

  1. Download the model folder, either from the Files tab of this page or with the Hugging Face CLI:

    hf download emass-ai/rock_paper_scissors_cnn --local-dir rock_paper_scissors_cnn
    
  2. In ECS-Studio (1.140.0 or later), open a project for the ECS-DoT EVB, open the ECS-Compiler panel and choose rock_paper_scissors_cnn.onnx as the model file. The panel loads rock_paper_scissors_cnn_train.npz (calibration) and rock_paper_scissors_cnn_test.npz (evaluation) from the same folder automatically.

  3. Click Start Pipeline, then Build and Upload to run the model on the board. The ECS-Compiler User Guide describes each step.

Model Details

Model type image_classification
Architecture Conv9x9 s2 1->16 (32->12) + BN + ReLU; Conv3x3 s2 p1 16->48 (12->6) + BN + ReLU; Flatten(1728); Linear 1728->3; Softmax
Input input [1, 1, 32, 32] float32 (greyscale hand image, 1x32x32 (NCHW))
Preprocessing 300x300 RGB -> luma (PIL convert L) -> 32x32 box (area) downscale -> x = (pixel/255 - 0.4441) / 0.2421 (exact constants in training/vision/rps_train.py output); layout NCHW [1,1,32,32]
Output output [1, 3] float32
Labels paper, rock, scissors
Model size (float) 53.5 KB
Parameters 13,459

Performance Summary

Model Precision Board Inference Time Library Size Compute Unit
rock_paper_scissors_cnn int8 ECS-DoT EVB 4.44 ms 37.9 KB DLA + CPU

Inference time is the measured time of one dla_infer() call on the board (median of 204 runs, ECS-DoT at 50 MHz). Library size is the compiled model library libdla.a (5 layers on the DLA, 1 on the CPU).

On-board verification: Run on 3 real samples. The board's output matched the compiler's expected output exactly, and its prediction matched ONNX Runtime on every sample (cosine similarity 1.0000 or higher). On the 200-sample evaluation set the board (int8) scored 100.00% against 100.00% for the float model on the same samples, and agreed with ONNX Runtime on 200 of 200 samples.

Accuracy

100.00% top-1 accuracy on the Rock Paper Scissors dataset, official test split (372 images, CC BY 2.0), measured with ONNX Runtime on the float model. The images are computer-generated hands on a plain white background, so this is an easy benchmark; real photographs with cluttered backgrounds will need more training data.

Training Data

Files

File Contents Download
rock_paper_scissors_cnn.onnx The ONNX model Download
rock_paper_scissors_cnn_train.npz Calibration set the board build quantizes with Download
rock_paper_scissors_cnn_test.npz Evaluation set Download
sample_input.npz The real samples run on the board (x, with labels y) Download
LICENSE License Download
README.md This model card Download

Intended Use

  • This model is provided to evaluate and validate ECS-Compiler and the ECS-DoT. It is not a product and has not been tested for any particular end use.

License

  • This model is released under apache-2.0; see LICENSE. The training data keeps its own licence (see Training Data).

Provenance

  • Verified with ECS-Studio 1.140.0 (ECS-Compiler engine 0.6.0) on the ECS-DoT EVB, 2026-11-17. Build fingerprint d39e6394a000.
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support