MobileNetv2

README.md

@@ -5,11 +5,12 @@ examples constructed using the [Burn](https://github.com/burn-rs/burn) deep lear
 
 ## Collection of Official Models
 
-| Model                                          | Description                                           | Repository Link                              |
-|------------------------------------------------|-------------------------------------------------------|----------------------------------------------|
-| [SqueezeNet](https://arxiv.org/abs/1602.07360) | A small CNN-based model for image classification.     | [squeezenet-burn](squeezenet-burn/README.md) |
-| [ResNet](https://arxiv.org/abs/1512.03385)     | A CNN based on residual blocks with skip connections. | [resnet-burn](resnet-burn/README.md)         |
-| [RoBERTa](https://arxiv.org/abs/1907.11692)    | A robustly optimized BERT pretraining approach.       | [bert-burn](bert-burn/README.md)             |
+| Model                                           | Description                                           | Repository Link                              |
+|-------------------------------------------------|-------------------------------------------------------|----------------------------------------------|
+| [MobileNetV2](https://arxiv.org/abs/1801.04381) | A CNN model targeted at mobile devices.               | [mobilenetv2-burn](mobilenetv2-burn/README.md) |
+| [SqueezeNet](https://arxiv.org/abs/1602.07360)  | A small CNN-based model for image classification.     | [squeezenet-burn](squeezenet-burn/README.md) |
+| [ResNet](https://arxiv.org/abs/1512.03385)      | A CNN based on residual blocks with skip connections. | [resnet-burn](resnet-burn/README.md)         |
+| [RoBERTa](https://arxiv.org/abs/1907.11692)     | A robustly optimized BERT pretraining approach.       | [bert-burn](bert-burn/README.md)             |
 
 ## Community Contributions
 

mobilenetv2-burn/Cargo.toml

@@ -0,0 +1,25 @@
+[package]
+authors = ["Arjun31415", "guillaumelagrange <[email protected]>"]
+license = "MIT OR Apache-2.0"
+name = "mobilenetv2-burn"
+version = "0.1.0"
+edition = "2021"
+
+[features]
+default = []
+std = []
+pretrained = ["burn/network", "std", "dep:dirs"]
+
+[dependencies]
+# Note: default-features = false is needed to disable std
+burn = { version = "0.13.0" }
+burn-import = { version = "0.13.0" }
+dirs = { version = "5.0.1", optional = true }
+serde = { version = "1.0.192", default-features = false, features = [
+  "derive",
+  "alloc",
+] } # alloc is for no_std, derive is needed
+
+[dev-dependencies]
+burn = { version = "0.13.0", features = ["ndarray"] }
+image = { version = "0.24.9", features = ["png", "jpeg"] }

mobilenetv2-burn/NOTICES.md

@@ -0,0 +1,16 @@
+# NOTICES AND INFORMATION
+
+This file contains notices and information required by libraries that this repository copied or derived from. The use of the following resources complies with the licenses provided.
+
+## Sample Image
+
+Image Title: Standing yellow Labrador Retriever dog.
+Author: Djmirko
+Source: https://commons.wikimedia.org/wiki/File:YellowLabradorLooking_new.jpg
+License: https://creativecommons.org/licenses/by-sa/3.0/
+
+## Pre-trained Model
+
+The ImageNet pre-trained model was ported from [`torchvision.models.MobileNet_V2_Weights.IMAGENET1K_V2`](https://pytorch.org/vision/main/models/generated/torchvision.models.mobilenet_v2.html#torchvision.models.MobileNet_V2_Weights).
+
+As opposed to [other pre-trained models](https://pytorch.org/vision/stable/models/generated/torchvision.models.regnet_y_128gf.html#torchvision.models.RegNet_Y_128GF_Weights) in `torchvision`, no specific license was linked to the weights, which are assumed to be under the library's [BSD-3-Clause license](https://github.com/pytorch/vision/blob/main/LICENSE) ([ref](https://github.com/pytorch/vision/issues/160)).

mobilenetv2-burn/README.md

@@ -0,0 +1,40 @@
+# MobileNetV2 Burn
+
+[MobileNetV2](https://arxiv.org/abs/1801.04381) is a convolutional neural network architecture for
+classification tasks which seeks to perform well on mobile devices. You can find the
+[Burn](https://github.com/tracel-ai/burn) implementation for the MobileNetV2 in
+[src/model/mobilenetv2.rs](src/model/mobilenetv2.rs).
+
+The model is [no_std compatible](https://docs.rust-embedded.org/book/intro/no-std.html).
+
+## Usage
+
+### `Cargo.toml`
+
+Add this to your `Cargo.toml`:
+
+```toml
+[dependencies]
+mobilenetv2-burn = { git = "https://github.com/tracel-ai/models", package = "mobilenetv2-burn", default-features = false }
+```
+
+If you want to get the pre-trained ImageNet weights, enable the `pretrained` feature flag.
+
+```toml
+[dependencies]
+mobilenetv2-burn = { git = "https://github.com/tracel-ai/models", package = "mobilenetv2-burn", features = ["pretrained"] }
+```
+
+**Important:** this feature requires `std`.
+
+### Example Usage
+
+The [inference example](examples/inference.rs) initializes a MobileNetV2 from the ImageNet
+[pre-trained weights](https://pytorch.org/vision/main/models/generated/torchvision.models.mobilenet_v2.html#torchvision.models.MobileNet_V2_Weights)
+with the `NdArray` backend and performs inference on the provided input image.
+
+You can run the example with the following command:
+
+```sh
+cargo run --release --features pretrained --example inference samples/dog.jpg
+```

mobilenetv2-burn/examples/inference.rs

@@ -0,0 +1,69 @@
+use mobilenetv2_burn::model::{imagenet, mobilenetv2::MobileNetV2, weights};
+
+use burn::{
+    backend::NdArray,
+    tensor::{backend::Backend, Data, Device, Element, Shape, Tensor},
+};
+
+const HEIGHT: usize = 224;
+const WIDTH: usize = 224;
+
+fn to_tensor<B: Backend, T: Element>(
+    data: Vec<T>,
+    shape: [usize; 3],
+    device: &Device<B>,
+) -> Tensor<B, 3> {
+    Tensor::<B, 3>::from_data(Data::new(data, Shape::new(shape)).convert(), device)
+        // [H, W, C] -> [C, H, W]
+        .permute([2, 0, 1])
+        / 255 // normalize between [0, 1]
+}
+
+pub fn main() {
+    // Parse arguments
+    let img_path = std::env::args().nth(1).expect("No image path provided");
+
+    // Create MobileNetV2
+    let device = Default::default();
+    let model: MobileNetV2<NdArray> =
+        MobileNetV2::pretrained(weights::MobileNetV2::ImageNet1kV2, &device)
+            .map_err(|err| format!("Failed to load pre-trained weights.\nError: {err}"))
+            .unwrap();
+
+    // Load image
+    let img = image::open(&img_path)
+        .map_err(|err| format!("Failed to load image {img_path}.\nError: {err}"))
+        .unwrap();
+
+    // Resize to 224x224
+    let resized_img = img.resize_exact(
+        WIDTH as u32,
+        HEIGHT as u32,
+        image::imageops::FilterType::Triangle, // also known as bilinear in 2D
+    );
+
+    // Create tensor from image data
+    let img_tensor = to_tensor(
+        resized_img.into_rgb8().into_raw(),
+        [HEIGHT, WIDTH, 3],
+        &device,
+    )
+    .unsqueeze::<4>(); // [B, C, H, W]
+
+    // Normalize the image
+    let x = imagenet::Normalizer::new(&device).normalize(img_tensor);
+
+    // Forward pass
+    let out = model.forward(x);
+
+    // Output class index w/ score (raw)
+    let (score, idx) = out.max_dim_with_indices(1);
+    let idx = idx.into_scalar() as usize;
+
+    println!(
+        "Predicted: {}\nCategory Id: {}\nScore: {:.4}",
+        imagenet::CLASSES[idx],
+        idx,
+        score.into_scalar()
+    );
+}

mobilenetv2-burn/src/lib.rs

@@ -0,0 +1,3 @@
+#![cfg_attr(not(feature = "std"), no_std)]
+pub mod model;
+extern crate alloc;

mobilenetv2-burn/src/model/conv_norm.rs

@@ -0,0 +1,83 @@
+use burn::{
+    config::Config,
+    module::Module,
+    nn::{
+        conv::{Conv2d, Conv2dConfig},
+        BatchNorm, BatchNormConfig, PaddingConfig2d,
+    },
+    tensor::{self, backend::Backend, Tensor},
+};
+
+/// A rectified linear unit where the activation is limited to a maximum of 6.
+#[derive(Module, Debug, Clone, Default)]
+pub struct ReLU6 {}
+impl ReLU6 {
+    pub fn forward<B: Backend, const D: usize>(&self, input: Tensor<B, D>) -> Tensor<B, D> {
+        tensor::activation::relu(input).clamp_max(6)
+    }
+}
+
+/// A Conv2d -> BatchNorm -> activation block.
+#[derive(Module, Debug)]
+pub struct Conv2dNormActivation<B: Backend> {
+    conv: Conv2d<B>,
+    norm: BatchNorm<B, 2>,
+    activation: ReLU6,
+}
+
+/// [Conv2dNormActivation] configuration.
+#[derive(Config, Debug)]
+pub struct Conv2dNormActivationConfig {
+    pub in_channels: usize,
+    pub out_channels: usize,
+
+    #[config(default = "3")]
+    pub kernel_size: usize,
+
+    #[config(default = "1")]
+    pub stride: usize,
+
+    #[config(default = "None")]
+    pub padding: Option<usize>,
+
+    #[config(default = "1")]
+    pub groups: usize,
+
+    #[config(default = "1")]
+    pub dilation: usize,
+
+    #[config(default = false)]
+    pub bias: bool,
+}
+
+impl Conv2dNormActivationConfig {
+    pub fn init<B: Backend>(&self, device: &B::Device) -> Conv2dNormActivation<B> {
+        let padding = if let Some(padding) = self.padding {
+            padding
+        } else {
+            (self.kernel_size - 1) / 2 * self.dilation
+        };
+
+        Conv2dNormActivation {
+            conv: Conv2dConfig::new(
+                [self.in_channels, self.out_channels],
+                [self.kernel_size, self.kernel_size],
+            )
+            .with_padding(PaddingConfig2d::Explicit(padding, padding))
+            .with_stride([self.stride, self.stride])
+            .with_bias(self.bias)
+            .with_dilation([self.dilation, self.dilation])
+            .with_groups(self.groups)
+            .init(device),
+            norm: BatchNormConfig::new(self.out_channels).init(device),
+            activation: ReLU6 {},
+        }
+    }
+}
+impl<B: Backend> Conv2dNormActivation<B> {
+    pub fn forward(&self, input: Tensor<B, 4>) -> Tensor<B, 4> {
+        let x = self.conv.forward(input);
+        let x = self.norm.forward(x);
+        self.activation.forward(x)
+    }
+}