File size: 6,095 Bytes

f02d81b

---
license: apache-2.0
base_model: litert-community/RTMPose-Hand-LiteRT
base_model_relation: quantized
library_name: openvino
pipeline_tag: keypoint-detection
tags:
- openvino
- onnx
- intel-npu
---

# RTMPose-Hand-OpenVINO

RTMPose-Hand for Intel NPUs, GPUs and CPUs: rtmhand_fp16 from [litert-community/RTMPose-Hand-LiteRT](https://huggingface.co/litert-community/RTMPose-Hand-LiteRT/tree/df8c0805d04910fc7ff02800e43fdda38a415309) 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 1.77 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 |
|---|---|---|
| `rtmhand_fp16.onnx` | 55,523,602 | `d67dd75c36113cc9ce085b087a1cd7a23f65f30b0687b161cd4c708f040df136` |
| `openvino/rtmhand_fp16.xml` | 903,989 | `eacb32d5f1e6bd3c838652f48d50e2d02dd12769ee7db4dbe430dc04b7ab6f21` |
| `openvino/rtmhand_fp16.bin` | 55,054,332 | `bbb8a53c9bb02ba94821b741cb0c25d3c61d2f6e3adc6f1ab73c1b756da425ab` |
| `npu/rtmhand_fp16.blob` | 28,946,568 | `5e983dca709f92c023d49b4b7d592988c1bd234a768d2071bebb49edd105704c` |

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 to 2026-11-07. 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 (ONNX Runtime's CPU EP); PASS = cosine ≥ 0.99, no NaN or Inf, and top-1 agreement ≥ 0.99 for class scores. The Snapdragon rows are Qualcomm AI Hub profile jobs of the same ONNX files (their own timer).

## Speed and output check

| processor | runtime | file | ms | load s | output check |
|---|---|---|---|---|---|
| Intel AI Boost NPU | OpenVINO 2026.4.1 | `npu/` blob (`import_model`) | 1.77 | 0.02 | FAIL: output_0_output: argmax agreement 16/21; output_1_output: argmax agreement 20/21 |
| Intel AI Boost NPU | ONNX Runtime 1.31.1 + OpenVINO EP 1.8.1 | `.onnx`, compiled when the session loads | 1.99 | 1.33 | FAIL: output_0_output: argmax agreement 16/21; output_1_output: argmax agreement 20/21 |
| Intel Graphics (iGPU) | ONNX Runtime 1.31.1 + OpenVINO EP 1.8.1 | `.onnx` | 1.58 | 1.05 | PASS (cos ≥ 0.9999) |
| RTX 5080 Laptop (95 W limit) | ONNX Runtime 1.31.1, CUDA EP | fp16 copy of the `.onnx` (not in this repo) | 3.52 | 1.16 | FAIL: output_0_output: argmax agreement 20/21; output_1_output: argmax agreement 17/21 |
| RTX 5080 Laptop (95 W limit) | ONNX Runtime 1.31.1, CUDA EP | `.onnx` | 4.31 | 1.15 | PASS (cos ≥ 0.9999) |
| Core Ultra 9 386H CPU | ONNX Runtime 1.31.1, CPU EP | `.onnx` | 7.44 | 0.13 | reference |
| Snapdragon X2 Elite NPU | Qualcomm AI Hub, precompiled QNN ONNX | `.onnx` compiled by AI Hub | 0.96 | 0.59 | not compared (AI Hub profile job, 2026-10-30) |
| Snapdragon X2 Elite NPU | Qualcomm AI Hub, ONNX Runtime QNN EP | `.onnx`, compiled when it loads | 0.96 | 1.13 | not compared (AI Hub profile job, 2026-10-30) |
| Galaxy S26 NPU | Qualcomm AI Hub, precompiled QNN ONNX | `.onnx` compiled by AI Hub | 0.87 | 0.20 | not compared (AI Hub profile job, 2026-10-30) |

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:

```sh
pip install huggingface_hub
hf download mlboydaisuke/RTMPose-Hand-OpenVINO --local-dir RTMPose-Hand-OpenVINO
cd RTMPose-Hand-OpenVINO
```

OpenVINO (`pip install openvino==2026.4.1`):

```python
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/rtmhand_fp16.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/rtmhand_fp16.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 with the OpenVINO EP (`pip install onnxruntime==1.31.1 onnxruntime-ep-openvino==1.8.1`):

```python
import numpy as np
import onnxruntime as ort
import onnxruntime_ep_openvino as ov_ep

ort.register_execution_provider_library(ov_ep.get_ep_name(), ov_ep.get_library_path())
npu = [d for d in ort.get_ep_devices()
       if d.ep_name == ov_ep.get_ep_name() and d.ep_metadata.get("ov_device") == "NPU"]
if not npu:
    raise SystemExit("no Intel NPU visible to the OpenVINO EP")
so = ort.SessionOptions()
so.add_provider_for_devices(npu, {})
sess = ort.InferenceSession("rtmhand_fp16.onnx", sess_options=so)  # compiles for the NPU here
# replace the zeros with your preprocessed input (float32)
feeds = {i.name: np.zeros(i.shape, np.float32) for i in sess.get_inputs()}
outputs = sess.run(None, feeds)
print([o.shape for o in outputs])
```

## Licence

Apache-2.0, the licence of [litert-community/RTMPose-Hand-LiteRT](https://huggingface.co/litert-community/RTMPose-Hand-LiteRT). The source card names the upstream model as `open-mmlab/mmpose` (not a Hugging Face repo).