|
| 1 | +--- |
| 2 | +title: useObjectDetection |
| 3 | +sidebar_position: 1 |
| 4 | +--- |
| 5 | + |
| 6 | +Object detection is a computer vision technique that identifies and locates objects within images or video. It’s commonly used in applications like image recognition, video surveillance or autonomous driving. |
| 7 | +`useObjectDetection` is a hook that allows you to seamlessly integrate object detection into your React Native applications. |
| 8 | + |
| 9 | +:::caution |
| 10 | +It is recommended to use models provided by us, which are available at our [Hugging Face repository](https://huggingface.co/software-mansion/react-native-executorch-ssdlite320-mobilenet-v3-large). You can also use [constants](https://github.com/software-mansion/react-native-executorch/blob/69802ee1ca161d9df00def1dabe014d36341cfa9/src/constants/modelUrls.ts#L28) shipped with our library. |
| 11 | +::: |
| 12 | + |
| 13 | +## Reference |
| 14 | +```jsx |
| 15 | +import { useObjectDetection, SSDLITE_320_MOBILENET_V3_LARGE } from 'react-native-executorch'; |
| 16 | + |
| 17 | +function App() { |
| 18 | + const ssdlite = useObjectDetection({ |
| 19 | + modelSource: SSDLITE_320_MOBILENET_V3_LARGE, // alternatively, you can use require(...) |
| 20 | + }); |
| 21 | + |
| 22 | + ... |
| 23 | + for (const detection of await ssdlite.forward("https://url-to-image.jpg")) { |
| 24 | + console.log("Bounding box: ", detection.bbox); |
| 25 | + console.log("Bounding label: ", detection.label); |
| 26 | + console.log("Bounding score: ", detection.score); |
| 27 | + } |
| 28 | + ... |
| 29 | +} |
| 30 | +``` |
| 31 | + |
| 32 | +<details> |
| 33 | +<summary>Type definitions</summary> |
| 34 | + |
| 35 | +```typescript |
| 36 | +interface Bbox { |
| 37 | + x1: number; |
| 38 | + x2: number; |
| 39 | + y1: number; |
| 40 | + y2: number; |
| 41 | +} |
| 42 | + |
| 43 | +interface Detection { |
| 44 | + bbox: Bbox; |
| 45 | + label: keyof typeof CocoLabel; |
| 46 | + score: number; |
| 47 | +} |
| 48 | + |
| 49 | +interface ObjectDetectionModule { |
| 50 | + error: string | null; |
| 51 | + isReady: boolean; |
| 52 | + isGenerating: boolean; |
| 53 | + forward: (input: string) => Promise<Detection[]>; |
| 54 | +} |
| 55 | +``` |
| 56 | +</details> |
| 57 | + |
| 58 | +### Arguments |
| 59 | + |
| 60 | +`modelSource` |
| 61 | + |
| 62 | +A string that specifies the path to the model file. You can download the model from our [HuggingFace repository](https://huggingface.co/software-mansion/react-native-executorch-ssdlite320-mobilenet-v3-large/tree/main). |
| 63 | +For more information on that topic, you can check out the [Loading models](https://docs.swmansion.com/react-native-executorch/fundamentals/loading-models) page. |
| 64 | + |
| 65 | +### Returns |
| 66 | + |
| 67 | +The hook returns an object with the following properties: |
| 68 | + |
| 69 | + |
| 70 | +| **Field** | **Type** | **Description** | |
| 71 | +|-----------------------|---------------------------------------|------------------------------------------------------------------------------------------------------------------| |
| 72 | +| `forward` | `(input: string) => Promise<Detection[]>` | A function that accepts an image (url, b64) and returns an array of `Detection` objects. | |
| 73 | +| `error` | <code>string | null</code> | Contains the error message if the model loading failed. | |
| 74 | +| `isGenerating` | `boolean` | Indicates whether the model is currently processing an inference. | |
| 75 | +| `isReady` | `boolean` | Indicates whether the model has successfully loaded and is ready for inference. | |
| 76 | + |
| 77 | + |
| 78 | +## Running the model |
| 79 | + |
| 80 | +To run the model, you can use the `forward` method. It accepts one argument, which is the image. The image can be a remote URL, a local file URI, or a base64-encoded image. The function returns an array of `Detection` objects. Each object contains coordinates of the bounding box, the label of the detected object, and the confidence score. For more information, please refer to the reference or type definitions. |
| 81 | + |
| 82 | +## Detection object |
| 83 | +The detection object is specified as follows: |
| 84 | +```typescript |
| 85 | +interface Bbox { |
| 86 | + x1: number; |
| 87 | + y1: number; |
| 88 | + x2: number; |
| 89 | + y2: number; |
| 90 | +} |
| 91 | + |
| 92 | +interface Detection { |
| 93 | + bbox: Bbox; |
| 94 | + label: keyof typeof CocoLabels; |
| 95 | + score: number; |
| 96 | +} |
| 97 | +``` |
| 98 | +The `bbox` property contains information about the bounding box of detected objects. It is represented as two points: one at the bottom-left corner of the bounding box (`x1`, `y1`) and the other at the top-right corner (`x2`, `y2`). |
| 99 | +The `label` property contains the name of the detected object, which corresponds to one of the `CocoLabels`. The `score` represents the confidence score of the detected object. |
| 100 | + |
| 101 | + |
| 102 | + |
| 103 | +## Example |
| 104 | +```tsx |
| 105 | +import { useObjectDetection, SSDLITE_320_MOBILENET_V3_LARGE } from 'react-native-executorch'; |
| 106 | + |
| 107 | +function App() { |
| 108 | + const ssdlite = useObjectDetection({ |
| 109 | + modelSource: SSDLITE_320_MOBILENET_V3_LARGE, |
| 110 | + }); |
| 111 | + |
| 112 | + const runModel = async () => { |
| 113 | + const detections = await ssdlite.forward("https://url-to-image.jpg"); |
| 114 | + for (const detection of detections) { |
| 115 | + console.log("Bounding box: ", detection.bbox); |
| 116 | + console.log("Bounding label: ", detection.label); |
| 117 | + console.log("Bounding score: ", detection.score); |
| 118 | + } |
| 119 | + } |
| 120 | +} |
| 121 | +``` |
| 122 | + |
| 123 | +## Supported Models |
| 124 | + |
| 125 | +| Model | Number of classes | Class list | |
| 126 | +| --------------------------------------------------------------------------------------------------------------- | ----------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | |
| 127 | +| [SSDLite320 MobileNetV3 Large](https://pytorch.org/vision/main/models/generated/torchvision.models.detection.ssdlite320_mobilenet_v3_large.html#torchvision.models.detection.SSDLite320_MobileNet_V3_Large_Weights) | 91 | [COCO](https://github.com/software-mansion/react-native-executorch/blob/69802ee1ca161d9df00def1dabe014d36341cfa9/src/types/object_detection.ts#L14) | |
0 commit comments