Merge pull request #226 from benjaminrwilson/detection-eval-max-boxes…

…-fix

Limit max number of scores associated with max boxes.

argoverse/evaluation/detection/utils.py

@@ -157,9 +157,10 @@ def accumulate(
         logger.info(f"{dt_filtered.shape[0]} detections")
         logger.info(f"{gt_filtered.shape[0]} ground truth")
         if dt_filtered.shape[0] > 0:
-            ranked_detections, scores = rank(dt_filtered)
-            metrics = assign(ranked_detections, gt_filtered, cfg)
-            cls_to_accum[class_name] = np.hstack((metrics, scores))
+            ranked_dts, ranked_scores = rank(dt_filtered)
+
+            metrics = assign(ranked_dts, gt_filtered, cfg)
+            cls_to_accum[class_name] = np.hstack((metrics, ranked_scores[:, None]))
 
         cls_to_ninst[class_name] = gt_filtered.shape[0]
     return cls_to_accum, cls_to_ninst
@@ -212,10 +213,6 @@ def assign(dts: np.ndarray, gts: np.ndarray, cfg: DetectionCfg) -> np.ndarray:
             of true positive thresholds used for AP computation and S is the number of true positive errors.
     """
 
-    # Ensure the number of boxes considered per class is at most `MAX_NUM_BOXES`.
-    if dts.shape[0] > MAX_NUM_BOXES:
-        dts = dts[:MAX_NUM_BOXES]
-
     n_threshs = len(cfg.affinity_threshs)
     metrics = np.zeros((dts.shape[0], n_threshs + N_TP_ERRORS))
 
@@ -323,20 +320,28 @@ def filter_instances(
     return filtered_instances
 
 
-def rank(dts: List[ObjectLabelRecord]) -> Tuple[np.ndarray, np.ndarray]:
-    """Get the rankings for the detections, according to detector confidence.
+def rank(dts: np.ndarray) -> Tuple[np.ndarray, np.ndarray]:
+    """Rank the detections in descending order according to score (detector confidence).
+
 
     Args:
-        dts: Detections (N,).
+        dts: Array of `ObjectLabelRecord` objects. (N,).
 
     Returns:
-        ranks: Ranking for the detections (N,).
-        scores: Detection scores (N,).
+        ranked_dts: Array of `ObjectLabelRecord` objects ranked by score (N,) where N <= MAX_NUM_BOXES.
+        ranked_scores: Array of floats sorted in descending order (N,) where N <= MAX_NUM_BOXES.
     """
-    scores = np.array([dt.score for dt in dts])
+    scores = np.array([dt.score for dt in dts.tolist()])
     ranks = scores.argsort()[::-1]
-    ranked_detections = dts[ranks]
-    return ranked_detections, scores[:, np.newaxis]
+
+    ranked_dts = dts[ranks]
+    ranked_scores = scores[ranks]
+
+    # Ensure the number of boxes considered per class is at most `MAX_NUM_BOXES`.
+    if ranked_dts.shape[0] > MAX_NUM_BOXES:
+        ranked_dts = ranked_dts[:MAX_NUM_BOXES]
+        ranked_scores = ranked_scores[:MAX_NUM_BOXES]
+    return ranked_dts, ranked_scores
 
 
 def interp(prec: np.ndarray, method: InterpType = InterpType.ALL) -> np.ndarray: