Upload 37 files

src/components/MultiSourceCaptioningView.tsx

@@ -16,6 +16,24 @@ function parseFlatBoxArray(arr: any[]): { label: string, bbox_2d: number[] }[] {
   return [];
 }
 
+function normalizeBoxes(raw: any): { label: string, bbox_2d: number[] }[] {
+  let boxes = [];
+  if (Array.isArray(raw)) {
+    boxes = raw;
+  } else if (typeof raw === "object" && raw !== null) {
+    boxes = [raw];
+  }
+  // Normalize bbox_2d to flat [x1, y1, x2, y2]
+  return boxes.map(obj => {
+    let bbox = obj.bbox_2d;
+    if (Array.isArray(bbox) && Array.isArray(bbox[0])) {
+      // Convert [[x1, y1], [x2, y2]] to [x1, y1, x2, y2]
+      bbox = [bbox[0][0], bbox[0][1], bbox[1][0], bbox[1][1]];
+    }
+    return { ...obj, bbox_2d: bbox };
+  });
+}
+
 function isImageFile(file: File) {
   return file.type.startsWith("image/");
 }
@@ -64,6 +82,7 @@ export default function MultiSourceCaptioningView() {
       if (boxes.length === 0 && Array.isArray(output)) {
         boxes = parseFlatBoxArray(output);
       }
+      boxes = normalizeBoxes(boxes);
       if (boxes.length === 0) setInferenceStatus("No boxes detected or model output invalid.");
       drawBoundingBoxesOnCanvas(ctx, boxes);
     });
@@ -184,6 +203,7 @@ export default function MultiSourceCaptioningView() {
       if (boxes.length === 0 && Array.isArray(output)) {
         boxes = parseFlatBoxArray(output);
       }
+      boxes = normalizeBoxes(boxes);
       if (boxes.length === 0) setInferenceStatus("No boxes detected or model output invalid.");
       drawBoundingBoxesOnCanvas(ctx, boxes);
       setImageProcessed(true);