YOLACT-ResNet50-OpenVINO

YOLACT-ResNet50 for Intel NPUs, GPUs and CPUs: yolact from litert-community/YOLACT-ResNet50-LiteRT as ONNX (tf2onnx 1.17.0, opset 17), OpenVINO IR with fp32 weights, and a precompiled Intel NPU blob. On the NPU of a Core Ultra 9 386H the blob runs in 16.29 ms per run, numpy input to numpy output.

Scripts, raw records, and the same table for all 52 graphs of LiteRT-Models: https://github.com/john-rocky/windows-zoo

Files

file	bytes	sha256
yolact.onnx	124,705,334	414d1e9ad7c8e9075e0f285ae794307beaa5befe785b5fda22f1e0cc7cfb6eed
openvino/yolact.xml	789,193	4ff0eaf04fbf3f1f20531b03bf7bc799665aab73cc5872e14f1092e58206a6c3
openvino/yolact.bin	124,471,816	7d0014086613dd81ec686a5280679415df07d08b202883e4cbfe708007db2751
npu/yolact.blob	71,868,566	5a9dfa91a4a2a238c519439f48b795724890a4e15d5a0bb3e669958933f21961

The NPU blob was compiled with OpenVINO 2026.4.1, Intel NPU driver 32.0.100.4841 (OpenVINO reports NPU_DRIVER_VERSION 1004841, NPU_COMPILER_VERSION 524292). If import_model refuses a blob on another driver or OpenVINO version, compile the IR instead (the commented line in the snippet).

How it was measured

MSI Stealth 16 AI+ (Core Ultra 9 386H, RTX 5080 Laptop GPU at a 95 W power limit, Windows 11 build 26200.9711, on AC power), 2026-11-06. On the laptop, each row is one process: load, one run, 5 warmup runs, then the median of 30 timed runs (2 warmup + 10 timed runs when the first run takes more than 3 s), numpy input to numpy output, every runtime at its default settings, one process at a time, each started after the CPU, the RTX and the Intel GPU had gone idle (the thresholds are in METHOD.md of the GitHub repo). Output check: each output against the CPU output in fp32 on the same input (OpenVINO on the CPU, as ONNX Runtime 1.31.1 cannot load this file); PASS = cosine ≥ 0.99, no NaN or Inf, and top-1 agreement ≥ 0.99 for class scores.

Speed and output check

processor	runtime	file	ms	load s	output check
Intel AI Boost NPU	OpenVINO 2026.4.1	npu/ blob (import_model)	16.29	0.04	PASS (cos ≥ 0.9999)
Intel AI Boost NPU	ONNX Runtime 1.31.1 + OpenVINO EP 1.8.1	.onnx, compiled when the session loads	-	-	did not run
Intel Graphics (iGPU)	ONNX Runtime 1.31.1 + OpenVINO EP 1.8.1	.onnx	-	-	did not run
RTX 5080 Laptop (95 W limit)	ONNX Runtime 1.31.1, CUDA EP	fp16 copy of the .onnx (not in this repo)	-	-	did not run
RTX 5080 Laptop (95 W limit)	ONNX Runtime 1.31.1, CUDA EP	.onnx	-	-	did not run
Core Ultra 9 386H CPU	ONNX Runtime 1.31.1, CPU EP	.onnx	-	-	did not run
Core Ultra 9 386H CPU	OpenVINO 2026.4.1	.onnx, f32	-	-	reference (ONNX Runtime 1.31.1 cannot load this file)

The ONNX file against the source .tflite (LiteRT interpreter on the CPU, the same input): PASS, cosine ≥ 0.9999 for every output.

Run it on the Intel NPU

Download this repo, then run the snippets from its folder:

pip install huggingface_hub
hf download mlboydaisuke/YOLACT-ResNet50-OpenVINO --local-dir YOLACT-ResNet50-OpenVINO
cd YOLACT-ResNet50-OpenVINO

OpenVINO (pip install openvino==2026.4.1):

import numpy as np
import openvino as ov

core = ov.Core()
if "NPU" not in core.available_devices:
    raise SystemExit(f"no Intel NPU visible to OpenVINO: {core.available_devices}")
with open("npu/yolact.blob", "rb") as f:
    compiled = core.import_model(f.read(), "NPU")  # precompiled: no compile at load
# or compile the IR on this machine: compiled = core.compile_model("openvino/yolact.xml", "NPU")
# replace the zeros with your preprocessed input (float32, the shapes printed by compiled.inputs)
feeds = {i.get_any_name(): np.zeros(list(i.shape), np.float32) for i in compiled.inputs}
outputs = compiled(feeds)
print({o.get_any_name(): outputs[o].shape for o in compiled.outputs})

ONNX Runtime 1.31.1 with the OpenVINO EP 1.8.1 does not run yolact.onnx on the NPU: the session fails to load (onnxruntime.capi.onnxruntime_pybind11_state.Fail: [ONNXRuntimeError] : 1 : FAIL : Load model from C:\src\zoo\onnx\yolact.onnx failed:Node (__main__.Wrap/yolact.). The OpenVINO lines above run it.

Licence

MIT, the licence of litert-community/YOLACT-ResNet50-LiteRT. The upstream model: dbolya/yolact-resnet50.