Fix bugs in lyft eval and rename configs

configs/regnet/hv_pointpillars_regnet-400mf_secfpn_sbn-all_1x16_2x_lyft-3d.py → configs/regnet/hv_pointpillars_regnet-400mf_secfpn_sbn-all_2x8_2x_lyft-3d.py

-_base_ = './hv_pointpillars_regnet-400mf_fpn_sbn-all_1x16_2x_lyft-3d.py'
+_base_ = './hv_pointpillars_regnet-400mf_fpn_sbn-all_2x8_2x_lyft-3d.py'
 # model settings
 model = dict(
     pts_neck=dict(

mmdet3d/core/evaluation/lyft_eval.py

@@ -106,9 +106,7 @@ def lyft_eval(lyft, data_root, res_path, eval_set, output_dir, logger=None):
     predictions = load_lyft_predictions(res_path)
 
     class_names = get_class_names(gts)
-    print_log('Evaluating...', logger=logger)
-    class_table = AsciiTable([class_names], title='Class Names')
-    print_log(class_table.table, logger=logger)
+    print('Calculating mAP@0.5:0.95...')
 
     iou_thresholds = [0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85, 0.9, 0.95]
     metrics = {}
@@ -131,7 +129,8 @@ def lyft_eval(lyft, data_root, res_path, eval_set, output_dir, logger=None):
         row = [class_names[i], round(mAPs_cate[i], 3)]
         APs_data.append(row)
     APs_data.append(['Overall', round(final_mAP, 3)])
-    APs_table = AsciiTable(APs_data, title='mAPs@0.5:0.95')
+    APs_table = AsciiTable(
+        APs_data, title='mAPs@0.5:0.95', inner_footing_row_border=True)
     print_log(APs_table.table, logger=logger)
 
     res_path = osp.join(output_dir, 'lyft_metrics.json')

mmdet3d/datasets/lyft_dataset.py

@@ -299,7 +299,7 @@ class LyftDataset(Custom3DDataset):
         metric_prefix = f'{result_name}_Lyft'
 
         for i, name in enumerate(metrics['class_names']):
-            AP = float("f{round(metrics['mAPs_cate'][i], 3)}")
+            AP = float(metrics['mAPs_cate'][i])
             detail[f'{metric_prefix}/{name}_AP'] = AP
 
         detail[f'{metric_prefix}/mAP'] = metrics['Final mAP']