Update docs

benITo47 · benITo47 · commit 9b77da76a7f9 · 2026-03-20T14:42:19.000+01:00
diff --git a/docs/docs/02-benchmarks/inference-time.md b/docs/docs/02-benchmarks/inference-time.md
@@ -43,11 +43,21 @@ processing. Resizing is typically fast for small images but may be noticeably
 slower for very large images, which can increase total time.
 :::
 
+:::warning
+Times presented in the tables are measured for forward method with input size equal to 512. Other input sizes may yield slower or faster inference times.
+:::
+
 | Model / Device                                | iPhone 17 Pro [ms] | Google Pixel 10 [ms] |
 | :-------------------------------------------- | :----------------: | :------------------: |
 | SSDLITE_320_MOBILENET_V3_LARGE (XNNPACK FP32) |         20         |          18          |
 | SSDLITE_320_MOBILENET_V3_LARGE (Core ML FP32) |         18         |          -           |
 | SSDLITE_320_MOBILENET_V3_LARGE (Core ML FP16) |         8          |          -           |
+| RF_DETR_NANO (XNNPACK FP32)                   |        TBD         |         TBD          |
+| YOLO26N (XNNPACK FP32)                        |        TBD         |         TBD          |
+| YOLO26S (XNNPACK FP32)                        |        TBD         |         TBD          |
+| YOLO26M (XNNPACK FP32)                        |        TBD         |         TBD          |
+| YOLO26L (XNNPACK FP32)                        |        TBD         |         TBD          |
+| YOLO26X (XNNPACK FP32)                        |        TBD         |         TBD          |
 
 ## Style Transfer
 
diff --git a/docs/docs/02-benchmarks/memory-usage.md b/docs/docs/02-benchmarks/memory-usage.md
@@ -25,11 +25,21 @@ loaded and actively running inference, relative to the baseline app memory
 before model initialization.
 :::
 
+:::warning
+Data presented for YOLO models is based on inference with forward_640 method.
+:::
+
 | Model / Device                                | iPhone 17 Pro [MB] | Google Pixel 10 [MB] |
 | --------------------------------------------- | :----------------: | :------------------: |
 | SSDLITE_320_MOBILENET_V3_LARGE (XNNPACK FP32) |         94         |         104          |
 | SSDLITE_320_MOBILENET_V3_LARGE (Core ML FP32) |         83         |          -           |
 | SSDLITE_320_MOBILENET_V3_LARGE (Core ML FP16) |         62         |          -           |
+| RF_DETR_NANO (XNNPACK FP32)                   |        TBD         |         TBD          |
+| YOLO26N (XNNPACK FP32)                        |        TBD         |         TBD          |
+| YOLO26S (XNNPACK FP32)                        |        TBD         |         TBD          |
+| YOLO26M (XNNPACK FP32)                        |        TBD         |         TBD          |
+| YOLO26L (XNNPACK FP32)                        |        TBD         |         TBD          |
+| YOLO26X (XNNPACK FP32)                        |        TBD         |         TBD          |
 
 ## Style Transfer
 
diff --git a/docs/docs/02-benchmarks/model-size.md b/docs/docs/02-benchmarks/model-size.md
@@ -13,6 +13,12 @@ title: Model Size
 | Model                          | XNNPACK FP32 [MB] | Core ML FP32 [MB] | Core ML FP16 [MB] |
 | ------------------------------ | :---------------: | :---------------: | :---------------: |
 | SSDLITE_320_MOBILENET_V3_LARGE |       13.9        |       15.6        |       8.46        |
+| RF_DETR_NANO                   |        112        |         -         |         -         |
+| YOLO26N                        |       10.3        |         -         |         -         |
+| YOLO26S                        |       38.6        |         -         |         -         |
+| YOLO26M                        |       82.3        |         -         |         -         |
+| YOLO26L                        |        100        |         -         |         -         |
+| YOLO26X                        |        224        |         -         |         -         |
 
 ## Instance Segmentation
 
diff --git a/docs/docs/03-hooks/02-computer-vision/useObjectDetection.md b/docs/docs/03-hooks/02-computer-vision/useObjectDetection.md
@@ -61,13 +61,18 @@ You need more details? Check the following resources:
 - `error` - An error object if the model failed to load or encountered a runtime error.
 - `downloadProgress` - A value between 0 and 1 representing the download progress of the model binary.
 - `forward` - A function to run inference on an image.
+- `getAvailableInputSizes` - A function that returns available input sizes for multi-method models (YOLO). Returns `undefined` for single-method models.
 
 ## Running the model
 
 To run the model, use the [`forward`](../../06-api-reference/interfaces/ObjectDetectionType.md#forward) method. It accepts two arguments:
 
 - `input` (required) - The image to process. Can be a remote URL, a local file URI, a base64-encoded image (whole URI or only raw base64), or a [`PixelData`](../../06-api-reference/interfaces/PixelData.md) object (raw RGB pixel buffer).
-- `detectionThreshold` (optional) - A number between 0 and 1 representing the minimum confidence score for a detection to be included in the results. Defaults to `0.7`.
+- `options` (optional) - An [`ObjectDetectionOptions`](../../06-api-reference/interfaces/ObjectDetectionOptions.md) object with the following properties:
+  - `detectionThreshold` (optional) - A number between 0 and 1 representing the minimum confidence score. Defaults to model-specific value (typically `0.7`).
+  - `iouThreshold` (optional) - IoU threshold for non-maximum suppression (0-1). Defaults to model-specific value (typically `0.55`).
+  - `inputSize` (optional) - For multi-method models like YOLO, specify the input resolution (`384`, `512`, or `640`). Defaults to `384` for YOLO models.
+  - `classesOfInterest` (optional) - Array of class labels to filter detections. Only detections matching these classes will be returned.
 
 `forward` returns a promise resolving to an array of [`Detection`](../../06-api-reference/interfaces/Detection.md) objects, each containing:
 
@@ -78,11 +83,11 @@ To run the model, use the [`forward`](../../06-api-reference/interfaces/ObjectDe
 ## Example
 
 ```typescript
-import { useObjectDetection, RF_DETR_NANO } from 'react-native-executorch';
+import { useObjectDetection, YOLO26N } from 'react-native-executorch';
 
 function App() {
   const model = useObjectDetection({
-    model: RF_DETR_NANO,
+    model: YOLO26N,
   });
 
   const handleDetect = async () => {
@@ -91,13 +96,12 @@ function App() {
     const imageUri = 'file:///Users/.../photo.jpg';
 
     try {
-      const detections = await model.forward(imageUri, 0.5);
+      const detections = await model.forward(imageUri, {
+        detectionThreshold: 0.5,
+        inputSize: 640,
+      });
 
-      for (const detection of detections) {
-        console.log('Label:', detection.label);
-        console.log('Score:', detection.score);
-        console.log('Bounding box:', detection.bbox);
-      }
+      console.log('Detected:', detections.length, 'objects');
     } catch (error) {
       console.error(error);
     }
@@ -113,7 +117,15 @@ See the full guide: [VisionCamera Integration](./visioncamera-integration.md).
 
 ## Supported models
 
-| Model                                                                                                                         | Number of classes | Class list                                               |
-| ----------------------------------------------------------------------------------------------------------------------------- | ----------------- | -------------------------------------------------------- |
-| [SSDLite320 MobileNetV3 Large](https://huggingface.co/software-mansion/react-native-executorch-ssdlite320-mobilenet-v3-large) | 91                | [COCO](../../06-api-reference/enumerations/CocoLabel.md) |
-| [RF-DETR Nano](https://huggingface.co/software-mansion/react-native-executorch-rf-detr-nano)                                  | 80                | [COCO](../../06-api-reference/enumerations/CocoLabel.md) |
+| Model | Number of classes | Class list | Multi-size Support |\n| ----------------------------------------------------------------------------------------------------------------------------- | ----------------- | ------------------------------------------------------------ | ------------------ |
+| [SSDLite320 MobileNetV3 Large](https://huggingface.co/software-mansion/react-native-executorch-ssdlite320-mobilenet-v3-large) | 91 | [COCO](../../06-api-reference/enumerations/CocoLabel.md) | No |
+| [RF-DETR Nano](https://huggingface.co/software-mansion/react-native-executorch-rf-detr-nano) | 80 | [COCO](../../06-api-reference/enumerations/CocoLabel.md) | No |
+| [YOLO26N](https://huggingface.co/software-mansion/react-native-executorch-yolo26) | 80 | [COCO YOLO](../../06-api-reference/enumerations/CocoLabel.md) | Yes (384/512/640) |
+| [YOLO26S](https://huggingface.co/software-mansion/react-native-executorch-yolo26) | 80 | [COCO YOLO](../../06-api-reference/enumerations/CocoLabel.md) | Yes (384/512/640) |
+| [YOLO26M](https://huggingface.co/software-mansion/react-native-executorch-yolo26) | 80 | [COCO YOLO](../../06-api-reference/enumerations/CocoLabel.md) | Yes (384/512/640) |
+| [YOLO26L](https://huggingface.co/software-mansion/react-native-executorch-yolo26) | 80 | [COCO YOLO](../../06-api-reference/enumerations/CocoLabel.md) | Yes (384/512/640) |
+| [YOLO26X](https://huggingface.co/software-mansion/react-native-executorch-yolo26) | 80 | [COCO YOLO](../../06-api-reference/enumerations/CocoLabel.md) | Yes (384/512/640) |
+
+:::tip
+YOLO models support multiple input sizes (384px, 512px, 640px). Smaller sizes are faster but less accurate, while larger sizes are more accurate but slower. Choose based on your speed/accuracy requirements.
+:::
diff --git a/docs/docs/04-typescript-api/02-computer-vision/ObjectDetectionModule.md b/docs/docs/04-typescript-api/02-computer-vision/ObjectDetectionModule.md
@@ -43,12 +43,26 @@ For more information on loading resources, take a look at [loading models](../..
 To run the model, use the [`forward`](../../06-api-reference/classes/ObjectDetectionModule.md#forward) method. It accepts two arguments:
 
 - `input` (required) - The image to process. Can be a remote URL, a local file URI, a base64-encoded image (whole URI or only raw base64), or a [`PixelData`](../../06-api-reference/interfaces/PixelData.md) object (raw RGB pixel buffer).
-- `detectionThreshold` (optional) - A number between 0 and 1. Defaults to `0.7`.
+- `options` (optional) - An [`ObjectDetectionOptions`](../../06-api-reference/interfaces/ObjectDetectionOptions.md) object with:
+  - `detectionThreshold` (optional) - Minimum confidence score (0-1). Defaults to model-specific value.
+  - `iouThreshold` (optional) - IoU threshold for NMS (0-1). Defaults to model-specific value.
+  - `inputSize` (optional) - For YOLO models: `384`, `512`, or `640`. Defaults to `384`.
+  - `classesOfInterest` (optional) - Array of class labels to filter detections.
 
 The method returns a promise resolving to an array of [`Detection`](../../06-api-reference/interfaces/Detection.md) objects, each containing the bounding box, label, and confidence score.
 
 For real-time frame processing, use [`runOnFrame`](../../03-hooks/02-computer-vision/visioncamera-integration.md) instead.
 
+### Example with Options
+
+```typescript
+const detections = await model.forward(imageUri, {
+  detectionThreshold: 0.5,
+  inputSize: 640, // YOLO models only
+  classesOfInterest: ['PERSON', 'CAR'],
+});
+```
+
 ## Using a custom model
 
 Use [`fromCustomModel`](../../06-api-reference/classes/ObjectDetectionModule.md#fromcustommodel) to load your own exported model binary instead of a built-in preset.
