Create README.md

README.md

@@ -0,0 +1,136 @@
+---
+library_name: transformers.js
+tags:
+- pose-estimation
+license: apache-2.0
+---
+
+
+https://github.com/open-mmlab/mmpose/tree/main/projects/rtmo with ONNX weights to be compatible with Transformers.js.
+
+## Usage (Transformers.js)
+
+If you haven't already, you can install the [Transformers.js](https://huggingface.co/docs/transformers.js) JavaScript library from [NPM](https://www.npmjs.com/package/@xenova/transformers) using:
+```bash
+npm i @xenova/transformers
+```
+
+**Example:** Perform pose-estimation w/ `Xenova/RTMO-s`.
+
+```js
+import { AutoModel, AutoProcessor, RawImage } from '@xenova/transformers';
+
+// Load model and processor
+const model_id = 'Xenova/RTMO-s';
+const model = await AutoModel.from_pretrained(model_id);
+const processor = await AutoProcessor.from_pretrained(model_id);
+
+// Read image and run processor
+const url = 'https://huggingface.co/datasets/Xenova/transformers.js-docs/resolve/main/football-match.jpg';
+const image = await RawImage.read(url);
+const { pixel_values, original_sizes, reshaped_input_sizes } = await processor(image);
+
+// Predict bounding boxes and keypoints
+const { dets, keypoints } = await model({ input: pixel_values });
+
+// Select the first image
+const predicted_boxes = dets.tolist()[0];
+const predicted_points = keypoints.tolist()[0];
+const [height, width] = original_sizes[0];
+const [resized_height, resized_width] = reshaped_input_sizes[0];
+
+// Compute scale values
+const xScale = width / resized_width;
+const yScale = height / resized_height;
+
+// Define thresholds
+const point_threshold = 0.3;
+const box_threshold = 0.3;
+
+// Display results
+for (let i = 0; i < predicted_boxes.length; ++i) {
+    const [xmin, ymin, xmax, ymax, box_score] = predicted_boxes[i];
+    if (box_score < box_threshold) continue;
+
+    const x1 = (xmin * xScale).toFixed(2);
+    const y1 = (ymin * yScale).toFixed(2);
+    const x2 = (xmax * xScale).toFixed(2);
+    const y2 = (ymax * yScale).toFixed(2);
+
+    console.log(`Found person at [${x1}, ${y1}, ${x2}, ${y2}] with score ${box_score.toFixed(3)}`)
+    const points = predicted_points[i]; // of shape [17, 3]
+    for (let id = 0; id < points.length; ++id) {
+        const label = model.config.id2label[id];
+        const [x, y, point_score] = points[id];
+        if (point_score < point_threshold) continue;
+        console.log(`  - ${label}: (${(x * xScale).toFixed(2)}, ${(y * yScale).toFixed(2)}) with score ${point_score.toFixed(3)}`);
+    }
+}
+```
+
+<details>
+
+<summary>See example output</summary>
+
+```
+Found person at [423.33, 55.52, 644.28, 504.13] with score 0.988
+  - nose: (527.30, 117.12) with score 0.733
+  - left_eye: (541.79, 109.26) with score 0.554
+  - right_eye: (515.04, 107.59) with score 0.475
+  - left_shoulder: (563.30, 171.75) with score 1.000
+  - right_shoulder: (464.21, 159.75) with score 1.000
+  - left_elbow: (575.71, 238.04) with score 0.998
+  - right_elbow: (436.06, 218.10) with score 0.999
+  - left_wrist: (605.86, 303.35) with score 1.000
+  - right_wrist: (497.47, 220.82) with score 1.000
+  - left_hip: (540.97, 307.31) with score 1.000
+  - right_hip: (475.85, 318.78) with score 1.000
+  - left_knee: (578.63, 368.63) with score 1.000
+  - right_knee: (501.05, 442.49) with score 1.000
+  - left_ankle: (572.11, 464.96) with score 0.991
+  - right_ankle: (535.75, 441.52) with score 0.981
+Found person at [89.97, 3.96, 517.81, 507.28] with score 0.966
+  - left_shoulder: (242.65, 111.06) with score 0.999
+  - right_shoulder: (228.79, 112.54) with score 0.999
+  - left_elbow: (321.84, 169.07) with score 0.999
+  - right_elbow: (225.76, 218.20) with score 1.000
+  - left_wrist: (351.73, 220.74) with score 0.999
+  - right_wrist: (160.19, 228.03) with score 1.000
+  - left_hip: (342.34, 246.81) with score 1.000
+  - right_hip: (360.05, 259.35) with score 0.999
+  - left_knee: (299.56, 377.97) with score 0.998
+  - right_knee: (313.81, 378.83) with score 0.976
+  - left_ankle: (443.84, 312.35) with score 0.983
+  - right_ankle: (424.74, 433.61) with score 0.823
+Found person at [-0.53, 51.78, 153.65, 371.17] with score 0.769
+  - nose: (75.52, 85.67) with score 0.363
+  - left_shoulder: (121.54, 113.17) with score 1.000
+  - right_shoulder: (49.77, 117.60) with score 1.000
+  - left_elbow: (132.90, 147.02) with score 0.932
+  - right_elbow: (30.31, 156.42) with score 0.992
+  - left_wrist: (154.43, 162.08) with score 0.871
+  - right_wrist: (17.20, 196.43) with score 0.943
+  - left_hip: (105.61, 204.27) with score 0.999
+  - right_hip: (61.99, 203.66) with score 0.999
+  - left_knee: (114.70, 270.91) with score 1.000
+  - right_knee: (63.75, 275.33) with score 1.000
+  - left_ankle: (125.53, 342.00) with score 0.998
+  - right_ankle: (63.16, 344.07) with score 0.997
+Found person at [519.40, 34.94, 650.11, 312.07] with score 0.488
+  - nose: (554.82, 76.58) with score 0.920
+  - left_eye: (563.12, 69.41) with score 0.666
+  - right_eye: (544.82, 70.01) with score 0.595
+  - left_shoulder: (596.60, 105.61) with score 0.999
+  - right_shoulder: (523.29, 107.31) with score 0.969
+  - left_elbow: (625.14, 151.30) with score 0.999
+  - right_elbow: (515.96, 147.59) with score 0.322
+  - left_wrist: (630.90, 196.91) with score 0.998
+  - right_wrist: (520.75, 181.83) with score 0.415
+  - left_hip: (583.24, 200.84) with score 0.998
+  - right_hip: (533.69, 200.01) with score 0.978
+  - left_knee: (583.79, 265.14) with score 0.934
+  - right_knee: (538.27, 262.98) with score 0.669
+  - left_ankle: (584.90, 309.76) with score 0.489
+```
+
+</details>