fix eval script

README.md

@@ -288,11 +288,12 @@ python scripts/evaluate_semantic_segmentation.py \
 
 Should output:
 ```
-scene                                             iou     f1
-barsana_DJI_0500_0501_combined_sliced_2700_14700  63.36  75.32
-comana_DJI_0881_full                              60.55  73.75
-norway_210821_DJI_0015_full                       37.99  45.92
-overall avg                                       53.97  65.00
+scene                                             iou      f1                                                           
+barsana_DJI_0500_0501_combined_sliced_2700_14700  63.371  75.338
+comana_DJI_0881_full                              60.559  73.779
+norway_210821_DJI_0015_full                       37.986  45.939
+mean                                              53.972  65.019
+
 ```
 
 Not providing `--scenes` will make an average across all 3 scenes (not average after each metric individually):
@@ -308,18 +309,18 @@ all    60.456  73.261
 
 #### IoU
 
-| method |  barsana_DJI_0500_0501_combined_sliced_2700_14700 | comana_DJI_0881_full |  norway_210821_DJI_0015_full |
-|:-|:-|:-|:-|
-| [Mask2Former](https://openaccess.thecvf.com/content/CVPR2022/papers/Cheng_Masked-Attention_Mask_Transformer_for_Universal_Image_Segmentation_CVPR_2022_paper.pdf) | 63.36 | 60.55 | 37.99 |
-| [NGC(LR)](https://openaccess.thecvf.com/content/ICCV2023W/LIMIT/papers/Marcu_Self-Supervised_Hypergraphs_for_Learning_Multiple_World_Interpretations_ICCVW_2023_paper.pdf) | 46.51 | 45.59 | 30.17 |
-| [CShift](https://www.bmvc2021-virtualconference.com/assets/papers/0455.pdf)[^1] | 46.27 | 43.67 | 29.09 |
-| [NGC](https://cdn.aaai.org/ojs/16283/16283-13-19777-1-2-20210518.pdf)[^1] | 44.34 | 38.99 | 22.63 |
+| method | average | barsana_DJI_0500_0501_combined_sliced_2700_14700 | comana_DJI_0881_full |  norway_210821_DJI_0015_full |
+|:-|:-|:-|:-|:-|
+| [Mask2Former](https://openaccess.thecvf.com/content/CVPR2022/papers/Cheng_Masked-Attention_Mask_Transformer_for_Universal_Image_Segmentation_CVPR_2022_paper.pdf) | 53.97 | 63.37 | 60.55 | 37.98 |
+| [NGC(LR)](https://openaccess.thecvf.com/content/ICCV2023W/LIMIT/papers/Marcu_Self-Supervised_Hypergraphs_for_Learning_Multiple_World_Interpretations_ICCVW_2023_paper.pdf) | 40.75  | 46.51 | 45.59 | 30.17 |
+| [CShift](https://www.bmvc2021-virtualconference.com/assets/papers/0455.pdf)[^1] | 39.67 | 46.27 | 43.67 | 29.09 |
+| [NGC](https://cdn.aaai.org/ojs/16283/16283-13-19777-1-2-20210518.pdf)[^1] | 35.32 | 44.34 | 38.99 | 22.63 |
 
 [^1]: reported in the [Dronescapes paper](https://openaccess.thecvf.com/content/ICCV2023W/LIMIT/papers/Marcu_Self-Supervised_Hypergraphs_for_Learning_Multiple_World_Interpretations_ICCVW_2023_paper.pdf).
 
 #### F1 Score
 
-| method |  barsana_DJI_0500_0501_combined_sliced_2700_14700 | comana_DJI_0881_full |  norway_210821_DJI_0015_full |
-|:-|:-|:-|:-|
-| [Mask2Former](https://openaccess.thecvf.com/content/CVPR2022/papers/Cheng_Masked-Attention_Mask_Transformer_for_Universal_Image_Segmentation_CVPR_2022_paper.pdf) | 75.32 | 73.75 | 45.92 |
+| method | mean | barsana_DJI_0500_0501_combined_sliced_2700_14700 | comana_DJI_0881_full |  norway_210821_DJI_0015_full |
+|:-|:-|:-|:-|:-|
+| [Mask2Former](https://openaccess.thecvf.com/content/CVPR2022/papers/Cheng_Masked-Attention_Mask_Transformer_for_Universal_Image_Segmentation_CVPR_2022_paper.pdf) | 65.01 | 75.33 | 73.77 | 45.93 |
 

scripts/eval_script_old.py

@@ -15,16 +15,15 @@ done
 
 Then run this:
 cd /dronescapes/scripts
-python eval_script_old.py --gt_path ../data/test_set_annotated_only_per_scene/comana/semantic_segprop8/ --pred_path ../data/test_set_annotated_only_per_scene/comana/semantic_mask2former_swin_mapillary_converted/ --num_classes 8 -o results/comana --overwrite
-python eval_script_old.py --gt_path ../data/test_set_annotated_only_per_scene/barsana/semantic_segprop8/ --pred_path ../data/test_set_annotated_only_per_scene/barsana/semantic_mask2former_swin_mapillary_converted/ --num_classes 8 -o results/barsana --overwrite
-python eval_script_old.py --gt_path ../data/test_set_annotated_only_per_scene/norway/semantic_segprop8/ --pred_path ../data/test_set_annotated_only_per_scene/norway/semantic_mask2former_swin_mapillary_converted/ --num_classes 8 -o results/norway --overwrite
+python eval_script_old.py --gt_path ../data/test_set_annotated_only_per_scene/comana/semantic_segprop8/ --pred_path ../data/test_set_annotated_only_per_scene/comana/semantic_mask2former_swin_mapillary_converted/ --class_weights 0.28172092 0.30589653 0.13341699 0.05937348 0.00474491 0.05987466 0.08660721 0.06836531 --classes land forest residential road little-objects water sky hill -o results/comana --overwrite
+python eval_script_old.py --gt_path ../data/test_set_annotated_only_per_scene/barsana/semantic_segprop8/ --pred_path ../data/test_set_annotated_only_per_scene/barsana/semantic_mask2former_swin_mapillary_converted/ --class_weights 0.28172092 0.30589653 0.13341699 0.05937348 0.00474491 0.05987466 0.08660721 0.06836531 --classes land forest residential road little-objects water sky hill -o results/barsana --overwrite
+python eval_script_old.py --gt_path ../data/test_set_annotated_only_per_scene/norway/semantic_segprop8/ --pred_path ../data/test_set_annotated_only_per_scene/norway/semantic_mask2former_swin_mapillary_converted/ --class_weights 0.28172092 0.30589653 0.13341699 0.05937348 0.00474491 0.05987466 0.08660721 0.06836531 --classes land forest residential road little-objects water sky hill -o results/norway --overwrite
 """
 
 from __future__ import annotations
 import os
-import cv2
 import numpy as np
-import multiprocessing as mp
+import pandas as pd
 from natsort import natsorted
 from pathlib import Path
 import shutil
@@ -42,72 +41,48 @@ def convert_label2multi(label, class_id):
     return np.array(out, dtype=bool)
 
 def process_all_video_frames(gt_files: list[Path], pred_files: list[Path], class_id: int):
-    global_true_positives = 0
-    global_true_negatives = 0
-    global_false_positives = 0
-    global_false_negatives = 0
-    
+    TP, TN, FP, FN = {}, {}, {}, {}
     for gt_file, pred_file in tqdm(zip(gt_files, pred_files), total=len(gt_files), desc=f"{class_id=}"):
-        gt_label = np.load(gt_file, allow_pickle=True)["arr_0"]
-        net_label = np.load(pred_file, allow_pickle=True)["arr_0"]
-
-        if gt_label.shape == ():
-            gt_label = gt_label.item()['data']
-        gt_label = convert_label2multi(gt_label, class_id)
-        net_label = convert_label2multi(net_label, class_id)
+        gt_label_raw = np.load(gt_file, allow_pickle=True)["arr_0"]
+        net_label_raw = np.load(pred_file, allow_pickle=True)["arr_0"]
+        gt_label = convert_label2multi(gt_label_raw, class_id)
+        net_label = convert_label2multi(net_label_raw, class_id)
 
         true_positives = np.count_nonzero(gt_label * net_label)
         true_negatives = np.count_nonzero((gt_label + net_label) == 0)
         false_positives = np.count_nonzero((np.array(net_label, dtype=int) - np.array(gt_label, dtype=int)) > 0)
         false_negatives = np.count_nonzero((np.array(gt_label, dtype=int) - np.array(net_label, dtype=int)) > 0)
         
-        global_true_positives += true_positives
-        global_true_negatives += true_negatives
-        global_false_positives += false_positives
-        global_false_negatives += false_negatives
-        
-    global_precision = global_true_positives / (global_true_positives + global_false_positives + np.spacing(1))
-    global_recall = global_true_positives / (global_true_positives + global_false_negatives + np.spacing(1))
+        TP[gt_file.name] = true_positives
+        TN[gt_file.name] = true_negatives
+        FP[gt_file.name] = false_positives
+        FN[gt_file.name] = false_negatives
+    df = pd.DataFrame([TP, FP, TN, FN], index=["tp", "fp", "tn", "fn"]).T
+    global_TP, global_TN, global_FP, global_FN = sum(TP.values()), sum(TN.values()), sum(FP.values()), sum(FN.values())
+    global_precision = global_TP / (global_TP + global_FP + np.spacing(1))
+    global_recall = global_TP / (global_TP + global_FN + np.spacing(1))
     global_f1_score = (2 * global_precision * global_recall) / (global_precision + global_recall + np.spacing(1))
-    global_iou = global_true_positives / (global_true_positives + global_false_positives + global_false_negatives + np.spacing(1))
+    global_iou = global_TP / (global_TP + global_FP + global_FN + np.spacing(1))
     
     return (global_precision, global_recall, global_f1_score, global_iou)
 
 def join_results(args: argparse.Namespace):
-    assert args.num_classes in (7, 8, 10), args.num_classes
-    if args.num_classes == 7:
-        CLASS_NAMES = ['land', 'forest', 'residential', 'road', 'little-objects', 'water', 'sky']
-        CLASS_WEIGHTS = [0.28172092, 0.37426183, 0.13341699, 0.05937348, 0.00474491, 0.05987466, 0.08660721]
-        #[0.37426183 0.28172092 0.13341699 0.08660721 0.05987466 0.05937348 0.00474491]
-    elif args.num_classes == 8:
-        CLASS_NAMES = ['land', 'forest', 'residential', 'road', 'little-objects', 'water', 'sky', 'hill']
-        CLASS_WEIGHTS = [0.28172092, 0.30589653, 0.13341699, 0.05937348, 0.00474491, 0.05987466, 0.08660721, 0.06836531]
-        #[0.30589653 0.28172092 0.13341699 0.08660721 0.06836531 0.05987466 0.05937348 0.00474491]
-    elif args.num_classes == 10:
-        CLASS_NAMES = ['land', 'forest', 'low-level', 'road', 'high-level', 'cars', 'water', 'sky', 'hill', 'person']
-        CLASS_WEIGHTS = [0.28172092, 0.30589653, 0.09954808, 0.05937348, 0.03386891, 0.00445865, 0.05987466, 0.08660721, 0.06836531, 0.00028626]
-        # [0.30589653 0.28172092 0.09954808 0.08660721 0.06836531 0.05987466 0.05937348 0.03386891 0.00445865 0.00028626]
-
-
-    out_path = os.path.join(args.out_dir, 'joined_results_' + str(args.num_classes) + 'classes.txt')
+    out_path = os.path.join(args.out_dir, 'joined_results_' + str(len(args.classes)) + 'classes.txt')
     out_file = open(out_path, 'w')
 
     joined_f1_scores_mean = []
     joined_iou_scores_mean = []
 
-    for CLASS_ID in range(0, len(CLASS_NAMES)):
-        
+    for CLASS_ID in range(len(args.classes)):
         RESULT_FILE = os.path.join(args.out_dir, 'evaluation_dronescapes_CLASS_' + str(CLASS_ID) + '.txt')
-            
         result_file_lines = open(RESULT_FILE, 'r').read().splitlines()
-            
         for idx, line in enumerate(result_file_lines):
             if idx != 0:
                 splits = line.split(',')
                 f1_score = float(splits[2])
                 iou_score = float(splits[3])
                     
-        out_file.write('------------------------- ' + ' CLASS    ' + str(CLASS_ID) + ' - ' + CLASS_NAMES[CLASS_ID] + '    --------------------------------------------\n')
+        out_file.write('------------------------- ' + ' CLASS    ' + str(CLASS_ID) + ' - ' + args.classes[CLASS_ID] + '    --------------------------------------------\n')
         # F1Score
         out_file.write('F1-Score: ' + str(round(f1_score, 4)) + '\n')
         # Mean IOU
@@ -122,13 +97,33 @@ def join_results(args: argparse.Namespace):
     out_file.write('\n\n')  
 
     out_file.write('\n\n')
-    out_file.write('Weighted Mean F1-Score all classes: ' + str(round(np.sum(np.dot(joined_f1_scores_mean, CLASS_WEIGHTS)), 4)) + '\n')
-    out_file.write('Weighted Mean IOU all classes: ' + str(round(np.sum(np.dot(joined_iou_scores_mean, CLASS_WEIGHTS)), 4)) + '\n')
+    out_file.write('Weighted Mean F1-Score all classes: ' + str(round(np.sum(np.dot(joined_f1_scores_mean, args.class_weights)), 4)) + '\n')
+    out_file.write('Weighted Mean IOU all classes: ' + str(round(np.sum(np.dot(joined_iou_scores_mean, args.class_weights)), 4)) + '\n')
     out_file.write('\n\n')    
 
     out_file.close()
     print(f"Written to '{out_path}'")
 
+def compat_old_txt_file(args: Namespace):
+    (tempdir := Path(tempfile.TemporaryDirectory().name)).mkdir()
+    (tempdir / "gt").mkdir()
+    (tempdir / "pred").mkdir()
+    print(f"old pattern detected. Copying files to a temp dir: {tempdir}")
+    test_files = natsorted(open(args.txt_path, "r").read().splitlines())
+    scenes = natsorted(set(([os.path.dirname(x) for x in test_files])))
+    assert len(scenes) == 1, scenes
+    files = natsorted([x for x in test_files if scenes[0] in x])
+    gt_files = [f"{args.gt_path}/{f.split('/')[0]}/segprop{len(args.classes)}/{f.split('/')[1]}.npz" for f in files]
+    pred_files = [f"{args.pred_path}/{f.split('/')[0]}/{int(f.split('/')[1]):06}.npz" for f in files]
+    assert all(Path(x).exists() for x in [*gt_files, *pred_files])
+    for _file in gt_files:
+        os.symlink(_file, tempdir / "gt" / Path(_file).name)
+    for _file in pred_files:
+        os.symlink(_file, tempdir / "pred" / Path(_file).name)
+    args.gt_path = tempdir / "gt"
+    args.pred_path = tempdir / "pred"
+    args.txt_path = None
+
 def main(args: argparse.Namespace):
     gt_files = natsorted([x for x in args.gt_path.iterdir()], key=lambda x: Path(x).name)
     pred_files = natsorted([x for x in args.pred_path.iterdir()], key=lambda x: Path(x).name)
@@ -149,41 +144,38 @@ if __name__ == "__main__":
     Comana: /Date3/hpc/datasets/dronescapes/all_scenes/dataset_splits/20221208_new_comana_clip/only_manually_annotated_test_files_30.txt
     gt_path: /Date3/hpc/datasets/dronescapes/all_scenes
     pred_path/Date3/hpc/code/Mask2Former/demo_dronescapes/outputs_dronescapes_compatible/mapillary_sseg
+    NC = 7
+        CLASS_NAMES = ['land', 'forest', 'residential', 'road', 'little-objects', 'water', 'sky']
+        CLASS_WEIGHTS = [0.28172092, 0.37426183, 0.13341699, 0.05937348, 0.00474491, 0.05987466, 0.08660721]
+    NC = 8
+        CLASS_NAMES = ['land', 'forest', 'residential', 'road', 'little-objects', 'water', 'sky', 'hill']
+        CLASS_WEIGHTS = [0.28172092, 0.30589653, 0.13341699, 0.05937348, 0.00474491, 0.05987466, 0.08660721, 0.06836531]
+    NC = 10
+        CLASS_NAMES = ['land', 'forest', 'low-level', 'road', 'high-level', 'cars', 'water', 'sky', 'hill', 'person']
+        CLASS_WEIGHTS = [0.28172092, 0.30589653, 0.09954808, 0.05937348, 0.03386891, 0.00445865, 0.05987466, 0.08660721, 0.06836531, 0.00028626]
     """
     parser = argparse.ArgumentParser()
     parser.add_argument("--gt_path", type=Path, required=True)
     parser.add_argument("--pred_path", type=Path, required=True)
     parser.add_argument("--out_dir", "-o", required=True, type=Path, default=Path(__file__).parent / "out_dir")
-    parser.add_argument("--num_classes", type=int, default=8)
+    parser.add_argument("--classes", nargs="+")
+    parser.add_argument("--class_weights", type=float, nargs="+", required=True)
     parser.add_argument("--txt_path")
     parser.add_argument("--overwrite", action="store_true")
     args = parser.parse_args()
+    if args.classes is None:
+        print("Class names not provided")
+        args.classes = list(map(str, range(len(args.class_weights))))
+    assert len(args.classes) == len(args.class_weights), (args.classes, args.class_weights)
+    assert len(args.classes) in (7, 8, 10), len(args.classes)
     assert not args.out_dir.exists() or args.overwrite, f"'{args.out_dir}' exists. Use --overwrite"
     shutil.rmtree(args.out_dir, ignore_errors=True)
     os.makedirs(args.out_dir, exist_ok=True)
 
     if args.txt_path is not None:
-        (tempdir := Path(tempfile.TemporaryDirectory().name)).mkdir()
-        (tempdir / "gt").mkdir()
-        (tempdir / "pred").mkdir()
-        print(f"old pattern detected. Copying files to a temp dir: {tempdir}")
-
-        test_files = natsorted(open(args.txt_path, "r").read().splitlines())
-        scenes = natsorted(set(([os.path.dirname(x) for x in test_files])))
-        assert len(scenes) == 1, scenes
-        files = natsorted([x for x in test_files if scenes[0] in x])
-        gt_files = [f"{args.gt_path}/{f.split('/')[0]}/segprop{args.num_classes}/{f.split('/')[1]}.npz" for f in files]
-        pred_files = [f"{args.pred_path}/{f.split('/')[0]}/{int(f.split('/')[1]):06}.npz" for f in files]
-        assert all(Path(x).exists() for x in [*gt_files, *pred_files])
-        for _file in gt_files:
-            os.symlink(_file, tempdir / "gt" / Path(_file).name)
-        for _file in pred_files:
-            os.symlink(_file, tempdir / "pred" / Path(_file).name)
-        args.gt_path = tempdir / "gt"
-        args.pred_path = tempdir / "pred"
-        args.txt_path = None
-
-    for class_id in range(args.num_classes):
+        compat_old_txt_file(args)
+
+    for class_id in range(len(args.classes)):
         args.class_id = class_id
         main(args)
     join_results(args)

scripts/evaluate_semantic_segmentation.py

@@ -1,3 +1,4 @@
+#!/usr/bin/env python3
 """
 Evaluation script for semantic segmentation for dronescapes. Outputs F1Score and mIoU for the classes and each frame.
 Usage: ./evaluate_semantic_segmentation.py y_dir gt_dir --classes C1 .. Cn [--class_weights W1 .. Wn] -o results.csv
@@ -38,7 +39,8 @@ def compute_raw_stats_per_frame(reader: MultiTaskDataset, classes: list[str]) ->
     index = []
     for i in trange(len(reader)):
         x = reader[i]
-        y, gt = x[0]["pred"], x[0]["gt"]
+        y = x[0]["pred"].argmax(-1) if x[0]["pred"].dtype != tr.int64 else x[0]["pred"]
+        gt = x[0]["gt"].argmax(-1) if x[0]["gt"].dtype != tr.int64 else x[0]["gt"]
         res[i] = multiclass_stat_scores(y, gt, num_classes=len(classes), average=None)[:, 0:4]
         index.append(x[1])
     res = res.reshape(len(reader) * len(classes), 4)
@@ -89,13 +91,10 @@ def main(args: Namespace):
     assert (a := len(reader.all_files_per_repr["gt"])) == (b := len(reader.all_files_per_repr["pred"])), f"{a} vs {b}"
 
     # Compute TP, FP, TN, FN for each frame
-    if not args.output_path.exists():
-        raw_stats = compute_raw_stats_per_frame(reader, args.classes)
-        logger.info(f"Stored raw metrics file to: '{args.output_path}'")
-        raw_stats.to_csv(args.output_path)
-    else:
-        logger.info(f"Loading raw metrics from: '{args.output_path}'. Delete this file if you want to recompute.")
-        raw_stats = pd.read_csv(args.output_path, index_col=0)
+    raw_stats = compute_raw_stats_per_frame(reader, args.classes)
+    logger.info(f"Stored raw metrics file to: '{args.output_path}'")
+    Path(args.output_path).parent.mkdir(exist_ok=True, parents=True)
+    raw_stats.to_csv(args.output_path)
 
     # Compute Precision, Recall, F1, IoU for each class and put them together in the same df.
     metrics_per_class = pd.concat([compute_metrics_by_class(raw_stats, class_name) for class_name in args.classes])