Skip to content
imaged

Image classification

MobileNetV3-Small

The smallest credible ImageNet-1k classifier, around 10 MB on disk.

BSD-3-Clause
Explore model

About this model

MobileNetV3-Small returns top-K class probabilities over the 1000 ImageNet-1k classes. About 2.5M parameters, fp32 ONNX is around 10 MB. Use it when the inference budget is in single-digit milliseconds.

The build is a fresh torchvision IMAGENET1K_V1 export with normalize baked into the graph. On M1 Max CPU it warms to 16 ms per call, faster than CoreML on this size.

What you can build

  • Tag user photos with a coarse subject category before showing them.
  • Run on-device classification on phones with a 10 MB asset budget.
  • Build a quick visual filter without paying for a vision API.
  • Pre-screen frames to skip downstream models when nothing of interest is there.

Variants & downloads

Choose the build for your app

Model files are being prepared.

Explore the model’s examples, source links, and integration notes while download variants are added.

Performance

Results depend on the device, runtime, and test conditions. Each measurement keeps its own context.

M1 Max, macOS 15, ONNX Runtime 1.17

See original catalog hardware notes · CPU

16.3 ms

Warm inference

Catalog reference · variant unspecifiedCold start (ms): 20.58

Reference result from the existing catalog; conversion variant and measurement date were not recorded.

M1 Max, macOS 15, ONNX Runtime 1.17

See original catalog hardware notes · CoreML

18.2 ms

Warm inference

Catalog reference · variant unspecifiedCold start (ms): 62.52

Reference result from the existing catalog; conversion variant and measurement date were not recorded.

RTX 2060, Ubuntu 24.04, ONNX Runtime 1.24.3

See original catalog hardware notes · CUDA

2.8 ms

Warm inference

Catalog reference · variant unspecified

Reference result from the existing catalog; conversion variant and measurement date were not recorded.

Integration examples

Vision SDK integration

cpp

#include "imaged/sdk.hpp"
#include "imaged/families/classifier/classifier_family.hpp"

imaged::SDK::AI ai;
auto& config = ai.getOptions();
config.set("classifier.model", "weights/mobilenet_v3_small.onnx");
config.set("classifier.labels", "weights/labels.txt");
config.set("classifier.topK", 5);

imaged::SDK::ClassifierFamily family;
family.load(config);

imaged::SDK::Image image;
image.load("photo.jpg");

auto result = family.run(&image);
auto& classes = std::get<imaged::SDK::ClassificationResult>(result).classes;