---
title: 'MiliPoint Dataset: mmWave Radar Benchmark'
url: https://www.emergentmind.com/topics/milipoint-dataset
type: topic
---

# MiliPoint Dataset: mmWave Radar Benchmark

Searching arXiv for MiliPoint and directly related follow-up work.
MiliPoint is a large-scale, point-cloud–based mmWave radar dataset for human activity sensing that was introduced to provide a standardized benchmark for point-set deep learning on RF sensing data. It is designed around three core tasks—person identification, action classification, and human keypoint estimation—and uses sparse 3D point clouds derived from millimetre-wave radar rather than camera imagery, making the sensing modality intrinsically non-visual and less intrusive [2309.13425]. Subsequent work has also used MiliPoint as a testbed for privacy-aware human action recognition pipelines based on DBSCAN, the Hungarian algorithm, and Kalman filtering, emphasizing the dataset’s role in studying how classical point-cloud processing can stabilize sparse and noisy radar measurements for downstream deep models [2508.10469].

## 1. Definition and scope

MiliPoint is a mmWave radar point-cloud dataset specifically designed for human activity recognition and related sensing tasks. In the original dataset paper, it is presented as a large-scale, open dataset for the community to explore how mmWave radars can be utilised for human activity recognition, with point clouds used as the canonical data representation rather than image-like radar products such as range–Doppler heatmaps [2309.13425].

The dataset targets three tasks within a unified benchmark. The first is **identification**, in which the objective is to infer subject identity from stacked radar point clouds. The second is **action classification**, in which a temporal sequence of radar point clouds is mapped to an action label. The third is **keypoint estimation**, in which the model regresses 3D human keypoints in radar coordinates [2309.13425]. This joint coverage distinguishes MiliPoint from earlier mmWave datasets that predominantly targeted only keypoint or pose estimation, or only action classification.

A central characteristic of MiliPoint is that each radar frame is represented only by 3D coordinates \((x, y, z)\) of radar returns. The point clouds are sparse, noisy, and temporally unstable, reflecting the underlying sensing physics of mmWave RF reflections and the limitations of CFAR-based point extraction [2309.13425]. This makes the dataset simultaneously a benchmark for privacy-preserving sensing and for robust point-set learning under severe data sparsity.

## 2. Sensor platform and acquisition protocol

MiliPoint uses a **TI IWR1843** commercial mmWave radar operating in the **77–81 GHz** band, with **3 transmit (TX) and 4 receive (RX) antennas** in a time-division multiplexing MIMO configuration [2309.13425]. The reported radar signal configuration includes a **chirp duration** of \(T_{\text{chirp}} = 100\,\mu\text{s}\), a **chirp slope** of \(S = 40\,\text{MHz}/\mu\text{s}\), a **total bandwidth** of \(B = 4\,\text{GHz}\), an **ADC sampling rate** of \(f_s = 5\,\text{MHz}\), and **CFAR thresholds** of **10 dB** in both range and Doppler domains [2309.13425]. The range resolution is reported as about **4 cm**.

The acquisition setup also includes a **Zed 2 stereo camera**, used only to obtain ground-truth 3D human keypoints. The camera provides disparity and depth maps, and a pose-estimation network estimates 3D skeletons that are then transformed into the radar coordinate system via calibration [2309.13425]. The released dataset contains only radar point clouds and derived keypoints, while camera images are discarded after keypoint generation.

Data were collected in an **indoor environment**, controlled and largely free of external RF interference. Participants performed a **30-minute low-intensity cardio workout** consisting of **49 distinct actions**, with each action lasting approximately **30 s** and separated by **10 s breaks** [2309.13425]. Participants remained within a **1 m \(\times\) 1 m** square in front of the radar. The radar was placed **0.65 m** from the area center at a height of **1.0 m**, while the camera was placed **3.0 m** from the area center at a height of **0.7 m**. Radar point clouds and camera frames were both captured at **24 fps** and synchronized by timestamp [2309.13425].

This acquisition protocol establishes a controlled benchmark rather than a highly unconstrained naturalistic corpus. The dataset therefore emphasizes reproducibility and standardized evaluation over broad viewpoint or environmental variation.

## 3. Data representation, scale, and annotations

MiliPoint contains **545,059 radar frames**, or **545,000 frames** as summarized in later follow-up work, recorded from **11 participants** and covering **49 distinct human actions** [2309.13425] [2508.10469]. The participant cohort consists of **4 female** and **7 male** subjects, with average height \(171.84 \pm 10.41\) cm and average weight \(67.73 \pm 13.08\) kg; all participants had no mobility impairments [2309.13425].

Each radar frame is represented as a point cloud
\[
d \in \mathbb{R}^{N \times 3},
\]
where each row is a Cartesian coordinate \((x, y, z)\) corresponding to a detected reflection in metric 3D space [2309.13425]. Only the coordinate triplet is retained in the released representation; intensity, Doppler, or related radar attributes are not explicitly included in the dataset representation described in the paper.

Because the number of points \(N\) varies from frame to frame, the dataset uses a fixed-size representation. If a frame contains more than a maximum number of points \(k\), it is **randomly downsampled**; if it contains fewer than \(k\), it is **zero-padded**. Each frame is then represented as
\[
d \in \mathbb{R}^{k \times 3}.
\]
To encode motion, consecutive frames are stacked, yielding an input tensor
\[
d \in \mathbb{R}^{s \times k \times 3},
\]
where \(s\) is the stacking length [2309.13425].

The dataset provides labels for three tasks:

| Task | Label space | Notes |
|---|---|---|
| Identification | \(\mathbb{R}^{11}\) | 11 participant identities |
| Action classification | \(\mathbb{R}^{39}\) | 49 actions recorded; 39 used for classification |
| Keypoint estimation | \(\mathbb{R}^{9 \times 3}\) or \(\mathbb{R}^{18 \times 3}\) | Two keypoint granularities |

For keypoint estimation, MiliPoint defines both a **9-point** and an **18-point** setting. The 9-point setting includes Right Shoulder, Right Elbow, Left Shoulder, Left Elbow, Right Hip, Right Knee, Left Hip, Left Knee, and Head. The 18-point setting includes finer landmarks such as Nose, Neck, wrists, ankles, eyes, and ears, with Head excluded due to the finer facial landmark decomposition [2309.13425].

Recommended stacking hyperparameters are task-dependent: **\(s = 5\)** for identification, **\(s = 5\)** for both keypoint estimation tasks, and **\(s = 50\)** for action classification [2309.13425]. Follow-up work on preprocessing additionally states that temporal stacking is used so that each frame contains up to **1100 points**, following Cui et al. [2508.10469].

## 4. Benchmark tasks and baseline performance

The original benchmark uses an **80% training**, **10% validation**, and **10% test** split over samples, with evaluation performed separately for the three task families [2309.13425]. The principal metrics are **Top-1 accuracy** for identification, **Top-1** and **Top-3 accuracy** for action classification, and **Mean Localization Error (MLE)** in centimeters for keypoint estimation:
\[
\text{MLE} = \frac{1}{M}\sum_{i=1}^M \left\| \hat{\mathbf{p}}_i - \mathbf{p}_i \right\|_2.
\]
Here, \(\hat{\mathbf{p}}_i\) and \(\mathbf{p}_i\) denote predicted and ground-truth keypoint coordinates [2309.13425].

The baseline benchmark evaluates four representative point-based architectures: **DGCNN**, **Pointformer**, **PointNet++**, and **PointMLP** [2309.13425]. Their reported results are as follows.

| Model | Identification Top1 (%) | Action Top1 (%) | Action Top3 (%) |
|---|---:|---:|---:|
| DGCNN | \(77.65 \pm 0.92\) | \(13.61 \pm 2.09\) | \(34.59 \pm 2.74\) |
| Pointformer | \(83.94 \pm 0.81\) | \(29.27 \pm 0.55\) | \(50.44 \pm 1.18\) |
| PointNet++ | \(87.30 \pm 0.27\) | \(34.45 \pm 0.80\) | \(54.96 \pm 1.21\) |
| PointMLP | \(95.88 \pm 0.40\) | \(18.37 \pm 0.08\) | \(35.94 \pm 0.14\) |

For keypoint estimation, the corresponding MLE results are **16.53 ± 0.11 cm** and **18.51 ± 0.03 cm** for DGCNN, **14.99 ± 0.03 cm** and **17.03 ± 0.13 cm** for Pointformer, **13.55 ± 0.03 cm** and **14.94 ± 0.03 cm** for PointNet++, and **13.12 ± 0.30 cm** and **14.11 ± 0.22 cm** for PointMLP on the 9-point and 18-point settings respectively [2309.13425].

These results establish a characteristic task hierarchy. **Identification** is comparatively easy, with all models exceeding **75%** Top-1 accuracy and PointMLP reaching approximately **96%**. **Keypoint estimation** is also effective, with localization errors around **13–18 cm**. **Action classification** is substantially harder, with the best Top-1 result at only about **34–35%**, indicating that extracting action semantics from sparse mmWave point-cloud sequences remains far from solved [2309.13425].

This difficulty profile became especially prominent in later work, where training directly on raw MiliPoint point clouds yielded low action-recognition baselines—**13.61%** for DGCNN, **29.27%** for Pointformer, **34.45%** for PointNet++, and **18.37%** for PointMLP—before explicit preprocessing was applied [2508.10469].

## 5. Noise characteristics and preprocessing-oriented use

A defining property of MiliPoint is that the point clouds are **sparse** and **noisy**. The original dataset paper attributes this to thermal noise, CFAR thresholding, multipath reflections, clutter, and incomplete sampling of body surfaces, all of which produce variable and unstable point distributions across frames [2309.13425]. The later privacy-aware HAR study operationalizes this sparsity more concretely by describing a temporally stacked representation in which each frame contains up to **1100 points**, but many of these are zero-padding or non-informative detections [2508.10469].

In that follow-up pipeline, each frame is divided into **5 contiguous segments** of **220 points**. Null-value removal is performed using the threshold
\[
(p_x^2 + p_y^2 + p_z^2) > \text{threshold}^2,\quad \text{threshold}=0.001,
\]
which discards points near the origin treated as zero padding [2508.10469]. After this cleaning step, the average effective density becomes approximately **63 non-zero points per segment** in the 5-segment configuration, **132 non-zero points per segment** in a 2-segment configuration, and **32 non-zero points per segment** in a 10-segment configuration.

The same study evaluates three classical processing methods on MiliPoint: **DBSCAN**, the **Hungarian algorithm**, and **Kalman filtering** [2508.10469]. For DBSCAN, the selected parameters are \(\epsilon = 0.4\) and \(\text{min\_samples} = 6\), with a vertical weighting factor \(\alpha = 0.25\) that de-emphasizes the \(z\)-axis in the distance metric. Cluster centroids are computed as
\[
\text{Centroid} = \left(\frac{1}{n}\sum_{i=1}^{n} x_i,\; \frac{1}{n}\sum_{i=1}^{n} y_i,\; \frac{1}{n}\sum_{i=1}^{n} z_i\right),
\]
and the Hungarian algorithm then solves a centroid-assignment problem across segments using the Euclidean cost matrix
\[
C_{ij} = \| \text{centroid}_i^{(t)} - \text{centroid}_j^{(t+1)} \|_2.
\]

Kalman filtering is applied in the horizontal plane with state vector
\[
X = [\,x,\; y,\; v_x,\; v_y\,]^T
\]
and observation vector
\[
z = [x,\; y]^T,
\]
with parameters tuned by **Bayesian Optimization** to \(Q = 29.41\), \(R = 0.081\), and \(P = 14.64\), alongside a **2.0 meter** association threshold [2508.10469]. This pipeline reduces each sample from **1100 raw points** to approximately **60–300 points** in the final human cluster.

The study reports a representative four-track example in which Track 0 has **RMSE = 0.9317** and **median distance = 0.0964 m**, Track 1 has **RMSE = 1.3630** and **median distance = 1.4454 m**, Track 2 has **RMSE = 1.0531** and **median distance = 1.0316 m**, and Track 3 has **RMSE = 0.7158** and **median distance = 0.4745 m**; merging Tracks 0 and 3 yields the cluster judged both accurate and spatially aligned with the human ground truth [2508.10469].

This preprocessing-focused literature shows that MiliPoint functions not only as a benchmark for end-to-end point-set networks, but also as a benchmark for denoising, clustering, association, and tracking strategies tailored to sparse radar point clouds.

## 6. Comparison with earlier mmWave datasets and distinctive contributions

The original paper explicitly positions MiliPoint against prior mmWave datasets such as **mmPose**, **MARS**, **HuPR**, **mRI**, **CubeLearn**, and **RadHAR** [2309.13425]. In that comparison, the principal axes are task coverage, number of participants, dataset size, and number of actions.

| Dataset | Tasks | Participants | Size (frames) | # Actions |
|---|---|---:|---:|---:|
| mmPose | K | 2 | 15k | 4 |
| MARS | K | 4 | 40k | 10 |
| HuPR | K | 6 | 141k | 3 |
| mRI | K | 20 | 160k | 12 |
| CubeLearn | A | 8 | 1k | 6 |
| RadHAR | A | 2 | 167k | 5 |
| MiliPoint | A, I, K | 11 | 545k | 49 |

Within this comparison, MiliPoint is described as the **largest mmWave HAR dataset to date**, with **545k radar frames in total**, **49 distinct actions**, and support for all three key HAR tasks in one dataset [2309.13425]. The paper further states that MiliPoint is **4.08× more actions** than the most diverse earlier dataset and about **3.26× larger** than the largest prior datasets in frame count.

The significance of this comparison is methodological rather than merely statistical. Earlier datasets generally focused on a single task, often pose estimation only or action classification only, and many used non-point-cloud representations. MiliPoint instead standardizes a **point-cloud** representation and supplies a common platform for evaluating GNN-based, transformer-based, hierarchical set-abstraction, and MLP-based point architectures under the same sensing setup [2309.13425].

A plausible implication is that MiliPoint helped shift mmWave HAR benchmarking toward point-set learning as a first-class problem, rather than treating radar outputs primarily as 2D image-like tensors. The follow-up preprocessing study reinforces this reading by showing that classical geometric processing can materially change downstream classification performance on the same dataset [2508.10469].

## 7. Privacy, limitations, and access

MiliPoint is repeatedly framed as a **privacy-preserving** alternative to camera-based sensing because mmWave radar captures only RF reflections as sparse 3D points and does not encode facial appearance, clothing, or other direct visual identifiers [2309.13425] [2508.10469]. The dataset paper states that participants were fully informed and signed consent forms, that camera images were used only to compute keypoints and discarded immediately thereafter, that the released data contain only radar point clouds and derived keypoints, and that the study was approved by the **University of Bristol Faculty of Engineering Research Ethics Committee (reference 12802)** [2309.13425].

The dataset nevertheless has several explicit limitations. It is based on a **single-radar indoor setup** in a **low-clutter environment**; participants generally **face the radar**; angular variation is limited; movements are largely **low-intensity cardio** emphasizing limb motion; and more complex postures such as sitting or lying are not represented [2309.13425]. The radar’s **3 TX / 4 RX** arrangement is optimized for automotive settings and provides stronger azimuth than elevation resolution, which may reduce sensitivity to actions involving critical height changes. The paper also notes that **multipath effects** are not fully addressed, and that experiments are primarily **line-of-sight** [2309.13425].

Later work adds a further practical limitation: strong performance on MiliPoint action recognition may depend on explicit preprocessing. On the reported benchmark, **Kalman filtering alone** yields **98.45%**, **96.11%**, **98.48%**, and **98.57%** accuracy across DGCNN, Pointformer, PointNet++, and PointMLP respectively; **DBSCAN alone** yields **98.19%**, **96.14%**, **98.54%**, and **98.42%**; **Hungarian + DBSCAN** remains similarly strong; while the full **DS + KM + HG** pipeline reaches **95.26%**, **91.45%**, **95.67%**, and **95.31%** but with a much higher preprocessing time of **0.2369 s** per sample [2508.10469]. This suggests that benchmark outcomes on MiliPoint can be highly sensitive to the front-end processing regime.

MiliPoint is available through the **GitHub repository** `https://github.com/yizzfz/MiliPoint/`, with raw data hosted on **Google Drive** at `https://drive.google.com/file/d/1rq8yyokrNhAGQryx7trpUqKenDnTI6Ky/` [2309.13425]. The dataset is released under the **MIT license**. The original repository includes dataset documentation, code for loading and preprocessing point clouds, baseline models, training scripts, and pre-trained weights [2309.13425]. The later preprocessing pipeline is also publicly released at `https://github.com/Maimunatunau/Human-Action-Recognition-HAR-using-mmWave-Radar` [2508.10469].

Taken together, these properties make MiliPoint a reference dataset for privacy-aware mmWave HAR research: large enough to support comparative benchmarking, technically standardized enough to support reproducibility, and challenging enough to expose open problems in action recognition, temporal modeling, denoising, and robust point-cloud learning [2309.13425].

Source: https://www.emergentmind.com/topics/milipoint-dataset