Skip to content
ModelListedExternal
Download

Autolens Mobilenetv4

Lightweight CNN baseline for 8-class vehicle body type classification.

Sourcehuggingface.co/morpeN1/autolens-mobilenetv4/tree/fad77449bb89defaabe04dbc6b3d43cff55e9c94
KindOtherVersionRevision fad7744

Model source

Kind
Other
File size
32.1–32.5 MB
Version
vRevision fad7744
Updated
Sep 28, 2026
Downloads
10
Source
Hugging Face

Source description

AutoLens MobileNetV4 Baseline

Lightweight CNN baseline for 8-class vehicle body type classification.

Model

  • Architecture: MobileNetV4 Conv Medium
  • Classes: 8 vehicle body types
  • Source run: baseline_0_mobilenetv4_20260509_103508
  • Source checkpoint: checkpoints/baseline_0_mobilenetv4_20260509_103508/best-22-0.8350.ckpt
  • Export format: model.safetensors plus metadata.json, with ONNX for deployment

Classes

  • SUV
  • VAN
  • STATION WAGON
  • MICRO
  • OPEN WHEEL / F1
  • SEDAN
  • HATCHBACK
  • PICK UP

Preprocessing

  • RGB input
  • Resize: 256
  • Center crop / model input: 224 x 224
  • Mean: [0.4429, 0.4354, 0.437]
  • Std: [0.2456, 0.2421, 0.2449]

Internal Test Results

Evaluated on held-out internal test set (3170 samples, no data leakage).

  • Accuracy: 0.8905
  • F1-macro: 0.8510
  • F1-weighted: 0.8912
  • Precision-macro: 0.8458
  • Recall-macro: 0.8592

Per-class Performance

ClassPrecisionRecallF1-scoreSupport
SUV0.82840.87240.8498670
VAN0.94950.94950.9495455
STATION WAGON0.71210.71210.712166
MICRO0.86360.86360.863622
OPEN WHEEL / F10.98630.98630.9863586
SEDAN0.88110.90430.8926836
HATCHBACK0.70500.73680.7206266
PICK UP0.86330.89220.8775269

Artifact sizes

  • model.safetensors: 32.517 MB
  • model.onnx: 32.12 MB

Files

  • model.safetensors — weights-only model artifact
  • metadata.json — architecture, classes, preprocessing, source run, and artifact metadata
  • model.onnx — ONNX Runtime inference artifact
  • model.simplified.onnx — optional simplified ONNX graph when available
  • size_report.json / size_check.json — artifact size and ONNX smoke evidence

Intended use

Educational demo and report evidence for classifying uploaded vehicle images into the 8 AutoLens body-type classes.

Limitations

The dataset is assembled from public/open sources and may contain domain bias. Similar body styles such as hatchback, station wagon, and sedan can be ambiguous. The merged raw dataset is not redistributed in this model repository.

Sources

1 source
Hugging FacePrimaryVersion fad77449bb89defaabe04dbc6b3d43cff55e9c94
huggingface.co/morpeN1/autolens-mobilenetv4/tree/fad77449bb89defaabe04dbc6b3d43cff55e9c94

Verified Sep 28

Model artifacts

3 artifacts

model.onnx

onnx · 32.1 MB · SHA-256 2400cc1aad8f…e743 · Hugging Face

Download

model.safetensors

safetensors · 32.5 MB · SHA-256 a0cc2631387d…e793 · Hugging Face

Download

model.simplified.onnx

onnx · 32.1 MB · SHA-256 909b601ae48f…e1b5 · Hugging Face

Download

Source excerpts

2 excerpts
ReadmeSep 28huggingface.co/morpeN1/autolens-mobilenetv4

--- library_name: onnxruntime pipeline_tag: image-classification tags: - computer-vision - image-classification - cars - onnx - safetensors license: other --- # AutoLens MobileNetV4 Baseline Lightweight CNN baseline for 8-class vehicle body type classification. ## Model - Architecture: MobileNetV4 Conv Medium - Classes: 8 vehicle body types - Source run: `baseline_0_mobilenetv4_20260509_103508` - Source checkpoint: `checkpoints/baseline_0_mobilenetv4_20260509_103508/best-22-0.8350.ckpt` - Export format: `model.safetensors` plus `metadata.json`, with ONNX for deployment ## Classes - SUV - VAN - STATION WAGON - MICRO - OPEN WHEEL / F1 - SEDAN - HATCHBACK - PICK UP ## Preprocessing - RGB input - Resize: 256 - Center crop / model input: 224 x 224 - Mean: `[0.4429, 0.4354, 0.437]` - Std: `[0.2456, 0.2421, 0.2449]` ## Internal Test Results Evaluated on held-out internal test set (3170 samples, no data leakage). - Accuracy: **0.8905** - F1-macro: **0.8510** - F1-weighted: 0.8912 - Precision-macro: 0.8458 - Recall-macro: 0.8592 ### Per-class Performance | Class | Precision | Recall | F1-score | Support | |-------|-----------|--------|----------|---------| | SUV | 0.8284 | 0.8724 | 0.8498 | 670 | | VAN | 0.9495 | 0.9495 | 0.9495 | 455 | | STATION WAGON | 0.7121 | 0.7121 | 0.7121 | 66 | | MICRO | 0.8636 | 0.8636 | 0.8636 | 22 | | OPEN WHEEL / F1 | 0.9863 | 0.9863 | 0.9863 | 586 | | SEDAN | 0.8811 | 0.9043 | 0.8926 | 836 | | HATCHBACK | 0.7050 | 0.7368 | 0.7206 | 266 | | PICK UP | 0.8633 | 0.8922 | 0.8775 | 269 | ## Artifact sizes - `model.safetensors`: 32.517 MB - `model.onnx`: 32.12 MB ## Files - `model.safetensors` — weights-only model artifact - `metadata.json` — architecture, classes, preprocessing, source run, and artifact metadata - `model.onnx` — ONNX Runtime inference artifact - `model.simplified.onnx` — optional simplified ONNX graph when available - `size_report.json` / `size_check.json` — artifact size and ONNX smoke evidence ## Intended use Educational demo and report evidence for classifying uploaded vehicle images into the 8 AutoLens body-type classes. ## Limitations The dataset is assembled from public/open sources and may contain domain bias. Similar body styles such as hatchback, station wagon, and sedan can be ambiguous. The merged raw dataset is not redistributed in this model repository.

NoteSep 28huggingface.co/morpeN1/autolens-mobilenetv4

Source context: 10 downloads · 0 likes · Pipeline image-classification · Library onnxruntime · Repo morpeN1/autolens-mobilenetv4