Figure 2.b-c : Showing qualitative WNet3D performance on additional datasets

Contents

Figure 2.b-c : Showing qualitative WNet3D performance on additional datasets#

Show that self-supervised model can perform well on additional datasets, without requiring any additional training.

import numpy as np
from tifffile import imread, imwrite
import sys
import numpy as np

import pyclesperanto_prototype as cle
from stardist.matching import matching_dataset
sys.path.append("../..")

from utils import *
from plots import *
print("Used GPU: ", cle.get_device())
show_params()
#################
SAVE_PLOTS_AS_PNG = False
SAVE_PLOTS_AS_SVG = True

Used GPU:  <Intel(R) UHD Graphics 620 on Platform: Intel(R) OpenCL (1 refs)>
Plot parameters (set in plots.py) : 
- COLORMAP : ████████
- DPI : 200
- Data path : C:\Users\Cyril\Desktop\Code\CELLSEG_BENCHMARK
- Font size : 20
- Title font size : 25.0
- Label font size : 20.0

%load_ext autoreload
%autoreload 2

Data#

data_path = DATA_PATH / "RESULTS/WNET OTHERS/"

# list all folders in the data path
folders = [x for x in data_path.iterdir() if x.is_dir()]
folders

[WindowsPath('C:/Users/Cyril/Desktop/Code/CELLSEG_BENCHMARK/RESULTS/WNET OTHERS/Mouse-Skull-Nuclei-CBG'),
 WindowsPath('C:/Users/Cyril/Desktop/Code/CELLSEG_BENCHMARK/RESULTS/WNET OTHERS/Platynereis-ISH-Nuclei-CBG'),
 WindowsPath('C:/Users/Cyril/Desktop/Code/CELLSEG_BENCHMARK/RESULTS/WNET OTHERS/Platynereis-Nuclei-CBG'),
 WindowsPath('C:/Users/Cyril/Desktop/Code/CELLSEG_BENCHMARK/RESULTS/WNET OTHERS/processed_instance_labels'),
 WindowsPath('C:/Users/Cyril/Desktop/Code/CELLSEG_BENCHMARK/RESULTS/WNET OTHERS/Seb cFOS')]

def get_predictions(path):
    return [imread(f) for f in path.glob("*.tif")]
#################
gt_folder = "labels"
mouse_skull_gt = get_predictions(folders[0] / gt_folder)[0]
platynereis_ISH_gt = get_predictions(folders[1] / gt_folder)[0]
platynereis_gt = get_predictions(folders[2] / gt_folder)[0]

prediction_folder = "pred"
mouse_skull_pred = get_predictions(folders[0] / prediction_folder)[0]
platynereis_ISH_pred = get_predictions(folders[1] / prediction_folder)[0]
platynereis_pred = get_predictions(folders[2] / prediction_folder)[0]

# get second channel of predictions
mouse_skull_pred = mouse_skull_pred[1]
platynereis_ISH_pred = platynereis_ISH_pred[1]
platynereis_pred = platynereis_pred[1]

# get validation set to estimate thresholds
mouse_skull_val = imread(folders[0] / "TEST/X2_left_WNet3D_pred_1.tif")[1] # take channel 1 of WNet prediction (0 is background)
mouse_skull_val_gt = imread(folders[0] / "TEST/Y2_left.tif")
###
platynereis_ISH_val = imread(folders[1] / "TEST/downsampled_cropped_X02_train_WNet3D_pred.tif")[1]
platynereis_ISH_val_gt = imread(folders[1] / "TEST/downsampled_cropped_Y02_train.tif")
###
platynereis_val = imread(folders[2] / "TEST/downsampled_cropped_dataset_hdf5_150_0_WNet3D_pred_1.tif")[1]
platynereis_val_gt = imread(folders[2] / "TEST/downsampled_cropped_mask_dataset_hdf5_150_0.tif")

Computations#

Threshold predictions#

GT_labels_val = [mouse_skull_val_gt, platynereis_ISH_val_gt, platynereis_val_gt]
predictions_val = [mouse_skull_val, platynereis_ISH_val, platynereis_val]

thresh = np.arange(0, 1, 0.05)
rows = []
for t in thresh:
    for i, (gt, pred) in enumerate(zip(GT_labels_val, predictions_val)):
        dices_row = {"Threshold": t, "Fold": i, "Dice": dice_coeff(
            np.where(gt > 0, 1, 0),
            np.where(pred > t, 1, 0)
            )}
        rows.append(dices_row)
        
dices_df = pd.DataFrame(rows)

sns.lineplot(data=dices_df, x="Threshold", y="Dice", hue="Fold", palette="tab10")
plt.title("Dice metric for different thresholds for WNet3D and GT")
plt.vlines([0.45, 0.55], 0, 1, colors="red", linestyles="dashed")
plt.show()

../_images/61c36aa98a1eb4882a3dbabac8f0db7243cad3b77202412ce03fee0c473c4f47.png

dices_df.groupby("Threshold").mean().sort_values("Dice", ascending=False).head(5)

	Fold	Dice
Threshold
0.50	1.0	0.717895
0.55	1.0	0.708049
0.45	1.0	0.631312
0.60	1.0	0.603701
0.65	1.0	0.455269

predictions = [mouse_skull_pred, platynereis_ISH_pred, platynereis_pred]
GT_labels = [mouse_skull_gt, platynereis_ISH_gt, platynereis_gt]

predictions_thresholded = []
thresholds = [0.45, 0.55, 0.55]
for i, pred in enumerate(predictions):
    predictions_thresholded.append(np.where(pred > thresholds[i], 1, 0))

mouse_skull_instance = np.array(
    cle.voronoi_otsu_labeling(predictions_thresholded[0], outline_sigma=1, spot_sigma=15)
)
platynereis_ISH_instance = np.array(
    cle.voronoi_otsu_labeling(predictions_thresholded[1], outline_sigma=0.5, spot_sigma=2)
)
platynereis_instance = np.array(
    cle.voronoi_otsu_labeling(predictions_thresholded[2], outline_sigma=0.5, spot_sigma=2.75)
)

Additional mouse skull postprocessing#

mouse_skull_instance = np.array(cle.closing_labels(mouse_skull_instance, radius=8))

def remap_image(image, new_min=1, new_max=100):
    min_val = image.min()
    max_val = image.max()
    return (image - min_val) / (max_val - min_val) * (new_max - new_min) + new_min

mouse_skull_remap = remap_image(mouse_skull_pred)
mouse_skull_instance = cle.merge_labels_with_border_intensity_within_range(
    image=mouse_skull_remap,
    labels=mouse_skull_instance.astype(np.int32), 
    minimum_intensity=35, 
    maximum_intensity=100
    )
mouse_skull_instance = np.array(mouse_skull_instance)

_generate_touch_mean_intensity_matrix.py (30): generate_touch_mean_intensity_matrix is supposed to work with images of integer type only.
Loss of information is possible when passing non-integer images.
_opencl_execute.py (281): overflow encountered in cast

# import napari
# viewer = napari.Viewer()
# viewer.add_image(predictions_thresholded[0], colormap="turbo")
# viewer.add_labels(mouse_skull_instance)
# viewer.add_labels(mouse_skull_instance, name="mouse_skull_instance_closed")
# viewer.add_image(mouse_skull_remap, name="mouse_skull_pred_remap", colormap="turbo")
# viewer.add_labels(mouse_skull_instance, name="mouse_skull_instance_closed_merged")

# Show the predictions and the instance segmentation
# import napari
# viewer = napari.Viewer()

# viewer.add_image(predictions_thresholded[0], name="mouse_skull_pred", colormap="turbo")
# viewer.add_labels(mouse_skull_instance, name="mouse_skull_instance")
# viewer.add_image(predictions_thresholded[1], name="platynereis_ISH_pred", colormap="turbo")
# viewer.add_labels(platynereis_ISH_instance, name="platynereis_ISH_instance")
# viewer.add_image(predictions_thresholded[2], name="platynereis_pred", colormap="turbo")
# viewer.add_labels(platynereis_instance, name="platynereis_instance")

Plots#

predictions = [
    mouse_skull_instance,
    platynereis_ISH_instance,
    platynereis_instance,
   ]
GT_labels = [
    mouse_skull_gt,
    platynereis_ISH_gt,
    platynereis_gt,
    ]
names = [
    "Mouse skull",
    "Platynereis ISH",
    "Platynereis",
    ]

# save instance labels
for pred, name in zip(predictions, names):
    save_path = data_path / "processed_instance_labels"
    save_path.mkdir(exist_ok=True)
    imwrite(save_path / f"{name}.tif", pred.astype(np.uint32))

taus = [0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9]


model_stats = []
names_stats = []

for i, p in enumerate(predictions):
    print(f"Validating on {names[i]}")
    stats = [matching_dataset(
        GT_labels[i], 
        p,
        thresh=t, 
        show_progress=False
        ) for t in taus]
    model_stats.append(stats)
    for t in taus:
        names_stats.append(names[i])
    # uncomment for ALL plots : 
    plot_performance(taus, stats, name=names[i])
    print("*"*20)

Validating on Mouse skull
********************
Validating on Platynereis ISH
********************
Validating on Platynereis
********************

../_images/4552d9210a49119b15ce7ab959405cba4973dd85a2eae5d9f4b730eff980d9f8.png

../_images/b421becf6911649ce9f0c30957c69cb8b3e837990e5e49c30797be8481eca1c9.png

../_images/c7f89105d73d46f2c87e3924418e26918aa5c2e1329f3716e0c5e43dfe371b41.png

dfs = [dataset_matching_stats_to_df(s) for s in model_stats]
df = pd.concat(dfs)
df["Dataset"] = names_stats
df

	criterion	fp	tp	fn	precision	recall	accuracy	f1	n_true	n_pred	mean_true_score	mean_matched_score	panoptic_quality	by_image	Dataset
thresh
0.1	iou	1096	4061	845	0.787473	0.827762	0.676608	0.807115	4906	5157	0.558399	0.674588	0.544470	False	Mouse skull
0.2	iou	1219	3938	968	0.763622	0.802691	0.642939	0.782669	4906	5157	0.555335	0.691842	0.541483	False	Mouse skull
0.3	iou	1319	3838	1068	0.744231	0.782307	0.616546	0.762794	4906	5157	0.550126	0.703209	0.536404	False	Mouse skull
0.4	iou	1507	3650	1256	0.707776	0.743987	0.569156	0.725430	4906	5157	0.536541	0.721169	0.523158	False	Mouse skull
0.5	iou	1952	3205	1701	0.621485	0.653282	0.467337	0.636987	4906	5157	0.495302	0.758175	0.482948	False	Mouse skull
0.6	iou	2468	2689	2217	0.521427	0.548104	0.364660	0.534433	4906	5157	0.437325	0.797887	0.426417	False	Mouse skull
0.7	iou	3006	2151	2755	0.417103	0.438443	0.271866	0.427507	4906	5157	0.365661	0.834001	0.356541	False	Mouse skull
0.8	iou	3670	1487	3419	0.288346	0.303098	0.173391	0.295538	4906	5157	0.263805	0.870361	0.257225	False	Mouse skull
0.9	iou	4789	368	4538	0.071359	0.075010	0.037958	0.073139	4906	5157	0.069084	0.920993	0.067361	False	Mouse skull
0.1	iou	532	2484	168	0.823607	0.936652	0.780151	0.876500	2652	3016	0.630515	0.673159	0.590023	False	Platynereis ISH
0.2	iou	590	2426	226	0.804377	0.914781	0.748304	0.856034	2652	3016	0.627286	0.685722	0.587002	False	Platynereis ISH
0.3	iou	652	2364	288	0.783820	0.891403	0.715496	0.834157	2652	3016	0.621627	0.697358	0.581706	False	Platynereis ISH
0.4	iou	776	2240	412	0.742706	0.844646	0.653442	0.790402	2652	3016	0.605176	0.716485	0.566312	False	Platynereis ISH
0.5	iou	949	2067	585	0.685345	0.779412	0.574007	0.729358	2652	3016	0.575968	0.738978	0.538980	False	Platynereis ISH
0.6	iou	1226	1790	862	0.593501	0.674962	0.461578	0.631616	2652	3016	0.518761	0.768578	0.485446	False	Platynereis ISH
0.7	iou	1622	1394	1258	0.462202	0.525641	0.326158	0.491884	2652	3016	0.421467	0.801816	0.394401	False	Platynereis ISH
0.8	iou	2311	705	1947	0.233753	0.265837	0.142051	0.248765	2652	3016	0.226469	0.851910	0.211925	False	Platynereis ISH
0.9	iou	2920	96	2556	0.031830	0.036199	0.017229	0.033874	2652	3016	0.033264	0.918915	0.031128	False	Platynereis ISH
0.1	iou	58	800	252	0.932401	0.760456	0.720721	0.837696	1052	858	0.525770	0.691387	0.579173	False	Platynereis
0.2	iou	86	772	280	0.899767	0.733840	0.678383	0.808377	1052	858	0.521990	0.711313	0.575009	False	Platynereis
0.3	iou	115	743	309	0.865967	0.706274	0.636675	0.778010	1052	858	0.514955	0.729115	0.567259	False	Platynereis
0.4	iou	152	706	346	0.822844	0.671103	0.586379	0.739267	1052	858	0.502527	0.748808	0.553569	False	Platynereis
0.5	iou	194	664	388	0.773893	0.631179	0.532905	0.695288	1052	858	0.484465	0.767556	0.533672	False	Platynereis
0.6	iou	269	589	463	0.686480	0.559886	0.445874	0.616754	1052	858	0.445128	0.795033	0.490340	False	Platynereis
0.7	iou	369	489	563	0.569930	0.464829	0.344124	0.512042	1052	858	0.383217	0.824425	0.422140	False	Platynereis
0.8	iou	535	323	729	0.376457	0.307034	0.203529	0.338220	1052	858	0.264275	0.860736	0.291118	False	Platynereis
0.9	iou	802	56	996	0.065268	0.053232	0.030205	0.058639	1052	858	0.048994	0.920388	0.053970	False	Platynereis

plot_stat_comparison(taus=taus, stats_list=model_stats, model_names=names, metric="IoU", plt_size=(9, 6))
if SAVE_PLOTS_AS_PNG:
    plt.savefig("f1_comparison.png")
if SAVE_PLOTS_AS_SVG:
    plt.savefig("f1_comparison.svg", bbox_inches='tight')

../_images/ef8e99021dfc20d792b411136d789259c640377937b482f8532606abb87d3634.png

Sanity check : Images only (no WNet)#

The aim here is to check that using the WNet does provide a benefit over using Otsu thresholding and Voronoi-based instance segmentation directly on the images.

mouse_skull_image = imread(folders[0] / "X1.tif")
platynereis_ISH_image = imread(folders[1] / "X01_cropped_downsampled.tif") 
platynereis_image = imread(folders[2] / "downsmapled_cropped_dataset_hdf5_100_0.tif")

mouse_skull_instance = np.array(
    cle.voronoi_otsu_labeling(mouse_skull_image, outline_sigma=1, spot_sigma=15)
)
platynereis_ISH_instance = np.array(
    cle.voronoi_otsu_labeling(platynereis_ISH_image, outline_sigma=0.5, spot_sigma=2)
)
platynereis_instance = np.array(
    cle.voronoi_otsu_labeling(platynereis_image, outline_sigma=0.5, spot_sigma=2.75)
)

mouse_skull_instance = np.array(cle.closing_labels(mouse_skull_instance, radius=8))

def remap_image(image, new_min=1, new_max=100):
    min_val = image.min()
    max_val = image.max()
    return (image - min_val) / (max_val - min_val) * (new_max - new_min) + new_min

mouse_skull_remap = remap_image(mouse_skull_image)
mouse_skull_instance = cle.merge_labels_with_border_intensity_within_range(
    image=mouse_skull_remap,
    labels=mouse_skull_instance.astype(np.int32), 
    minimum_intensity=35, 
    maximum_intensity=100
    )
mouse_skull_instance = np.array(mouse_skull_instance)

_generate_touch_mean_intensity_matrix.py (30): generate_touch_mean_intensity_matrix is supposed to work with images of integer type only.
Loss of information is possible when passing non-integer images.
_opencl_execute.py (281): overflow encountered in cast

predictions_images_only = [
    mouse_skull_instance,
    platynereis_ISH_instance,
    platynereis_instance,
]

taus = [0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9]


model_stats_images_only = []
names_stats = []

for i, p in enumerate(predictions_images_only):
    print(f"Validating on {names[i]}")
    stats = [matching_dataset(
        GT_labels[i], 
        p,
        thresh=t, 
        show_progress=False
        ) for t in taus]
    model_stats_images_only.append(stats)
    for t in taus:
        names_stats.append(names[i]+"- Image") 
    # uncomment for ALL plots : 
    plot_performance(taus, stats, name=names[i]+"- Image only")
    print("*"*20)

Validating on Mouse skull
********************
Validating on Platynereis ISH
********************
Validating on Platynereis
********************

../_images/f72651e45ede0f46407c8f3bdde88f080ef0a48d84dfd8bf5367cf9b9f981615.png

../_images/0baff91a9a66a90dafa9060fcecefa6da387e1ed3803201c6a2108ce98f0e6f5.png

../_images/e36b59fba42a1cde30af2a32b2683bde09cfe7cafb8537aeffda293226b3712b.png

dfs = [dataset_matching_stats_to_df(s) for s in model_stats_images_only]
df_im_only = pd.concat(dfs)
df_im_only["Dataset"] = names_stats
df_all = pd.concat([df, df_im_only])
df_all

	criterion	fp	tp	fn	precision	recall	accuracy	f1	n_true	n_pred	mean_true_score	mean_matched_score	panoptic_quality	by_image	Dataset
thresh
0.1	iou	1096	4061	845	0.787473	0.827762	0.676608	0.807115	4906	5157	0.558399	0.674588	0.544470	False	Mouse skull
0.2	iou	1219	3938	968	0.763622	0.802691	0.642939	0.782669	4906	5157	0.555335	0.691842	0.541483	False	Mouse skull
0.3	iou	1319	3838	1068	0.744231	0.782307	0.616546	0.762794	4906	5157	0.550126	0.703209	0.536404	False	Mouse skull
0.4	iou	1507	3650	1256	0.707776	0.743987	0.569156	0.725430	4906	5157	0.536541	0.721169	0.523158	False	Mouse skull
0.5	iou	1952	3205	1701	0.621485	0.653282	0.467337	0.636987	4906	5157	0.495302	0.758175	0.482948	False	Mouse skull
0.6	iou	2468	2689	2217	0.521427	0.548104	0.364660	0.534433	4906	5157	0.437325	0.797887	0.426417	False	Mouse skull
0.7	iou	3006	2151	2755	0.417103	0.438443	0.271866	0.427507	4906	5157	0.365661	0.834001	0.356541	False	Mouse skull
0.8	iou	3670	1487	3419	0.288346	0.303098	0.173391	0.295538	4906	5157	0.263805	0.870361	0.257225	False	Mouse skull
0.9	iou	4789	368	4538	0.071359	0.075010	0.037958	0.073139	4906	5157	0.069084	0.920993	0.067361	False	Mouse skull
0.1	iou	532	2484	168	0.823607	0.936652	0.780151	0.876500	2652	3016	0.630515	0.673159	0.590023	False	Platynereis ISH
0.2	iou	590	2426	226	0.804377	0.914781	0.748304	0.856034	2652	3016	0.627286	0.685722	0.587002	False	Platynereis ISH
0.3	iou	652	2364	288	0.783820	0.891403	0.715496	0.834157	2652	3016	0.621627	0.697358	0.581706	False	Platynereis ISH
0.4	iou	776	2240	412	0.742706	0.844646	0.653442	0.790402	2652	3016	0.605176	0.716485	0.566312	False	Platynereis ISH
0.5	iou	949	2067	585	0.685345	0.779412	0.574007	0.729358	2652	3016	0.575968	0.738978	0.538980	False	Platynereis ISH
0.6	iou	1226	1790	862	0.593501	0.674962	0.461578	0.631616	2652	3016	0.518761	0.768578	0.485446	False	Platynereis ISH
0.7	iou	1622	1394	1258	0.462202	0.525641	0.326158	0.491884	2652	3016	0.421467	0.801816	0.394401	False	Platynereis ISH
0.8	iou	2311	705	1947	0.233753	0.265837	0.142051	0.248765	2652	3016	0.226469	0.851910	0.211925	False	Platynereis ISH
0.9	iou	2920	96	2556	0.031830	0.036199	0.017229	0.033874	2652	3016	0.033264	0.918915	0.031128	False	Platynereis ISH
0.1	iou	58	800	252	0.932401	0.760456	0.720721	0.837696	1052	858	0.525770	0.691387	0.579173	False	Platynereis
0.2	iou	86	772	280	0.899767	0.733840	0.678383	0.808377	1052	858	0.521990	0.711313	0.575009	False	Platynereis
0.3	iou	115	743	309	0.865967	0.706274	0.636675	0.778010	1052	858	0.514955	0.729115	0.567259	False	Platynereis
0.4	iou	152	706	346	0.822844	0.671103	0.586379	0.739267	1052	858	0.502527	0.748808	0.553569	False	Platynereis
0.5	iou	194	664	388	0.773893	0.631179	0.532905	0.695288	1052	858	0.484465	0.767556	0.533672	False	Platynereis
0.6	iou	269	589	463	0.686480	0.559886	0.445874	0.616754	1052	858	0.445128	0.795033	0.490340	False	Platynereis
0.7	iou	369	489	563	0.569930	0.464829	0.344124	0.512042	1052	858	0.383217	0.824425	0.422140	False	Platynereis
0.8	iou	535	323	729	0.376457	0.307034	0.203529	0.338220	1052	858	0.264275	0.860736	0.291118	False	Platynereis
0.9	iou	802	56	996	0.065268	0.053232	0.030205	0.058639	1052	858	0.048994	0.920388	0.053970	False	Platynereis
0.1	iou	930	3079	1827	0.768022	0.627599	0.527587	0.690746	4906	4009	0.417881	0.665842	0.459927	False	Mouse skull- Image
0.2	iou	1042	2967	1939	0.740085	0.604770	0.498823	0.665620	4906	4009	0.415167	0.686488	0.456940	False	Mouse skull- Image
0.3	iou	1161	2848	2058	0.710402	0.580514	0.469425	0.638923	4906	4009	0.409135	0.704780	0.450301	False	Mouse skull- Image
0.4	iou	1338	2671	2235	0.666251	0.544435	0.427771	0.599215	4906	4009	0.396264	0.727845	0.436135	False	Mouse skull- Image
0.5	iou	1675	2334	2572	0.582190	0.475744	0.354657	0.523612	4906	4009	0.365215	0.767671	0.401962	False	Mouse skull- Image
0.6	iou	2052	1957	2949	0.488152	0.398899	0.281259	0.439035	4906	4009	0.323102	0.809983	0.355611	False	Mouse skull- Image
0.7	iou	2424	1585	3321	0.395360	0.323074	0.216235	0.355580	4906	4009	0.273659	0.847047	0.301193	False	Mouse skull- Image
0.8	iou	2835	1174	3732	0.292841	0.239299	0.151660	0.263376	4906	4009	0.210670	0.880362	0.231867	False	Mouse skull- Image
0.9	iou	3597	412	4494	0.102769	0.083979	0.048453	0.092428	4906	4009	0.077364	0.921231	0.085148	False	Mouse skull- Image
0.1	iou	383	2390	262	0.861882	0.901207	0.787479	0.881106	2652	2773	0.612524	0.679671	0.598862	False	Platynereis ISH- Image
0.2	iou	437	2336	316	0.842409	0.880845	0.756232	0.861198	2652	2773	0.609594	0.692056	0.595997	False	Platynereis ISH- Image
0.3	iou	491	2282	370	0.822935	0.860483	0.726058	0.841290	2652	2773	0.604571	0.702595	0.591087	False	Platynereis ISH- Image
0.4	iou	595	2178	474	0.785431	0.821267	0.670773	0.802949	2652	2773	0.590669	0.719217	0.577494	False	Platynereis ISH- Image
0.5	iou	784	1989	663	0.717274	0.750000	0.578871	0.733272	2652	2773	0.558883	0.745177	0.546417	False	Platynereis ISH- Image
0.6	iou	1016	1757	895	0.633610	0.662519	0.479008	0.647742	2652	2773	0.510833	0.771046	0.499439	False	Platynereis ISH- Image
0.7	iou	1406	1367	1285	0.492968	0.515460	0.336865	0.503963	2652	2773	0.414491	0.804119	0.405246	False	Platynereis ISH- Image
0.8	iou	2059	714	1938	0.257483	0.269231	0.151560	0.263226	2652	2773	0.229082	0.850875	0.223972	False	Platynereis ISH- Image
0.9	iou	2682	91	2561	0.032816	0.034314	0.017060	0.033548	2652	2773	0.031596	0.920800	0.030891	False	Platynereis ISH- Image
0.1	iou	58	758	294	0.928922	0.720532	0.682883	0.811563	1052	816	0.500484	0.694603	0.563714	False	Platynereis- Image
0.2	iou	82	734	318	0.899510	0.697719	0.647266	0.785867	1052	816	0.497073	0.712426	0.559872	False	Platynereis- Image
0.3	iou	114	702	350	0.860294	0.667300	0.602058	0.751606	1052	816	0.489355	0.733335	0.551179	False	Platynereis- Image
0.4	iou	148	668	384	0.818627	0.634981	0.556667	0.715203	1052	816	0.478119	0.752965	0.538523	False	Platynereis- Image
0.5	iou	194	622	430	0.762255	0.591255	0.499197	0.665953	1052	816	0.458364	0.775240	0.516273	False	Platynereis- Image
0.6	iou	267	549	503	0.672794	0.521863	0.416224	0.587794	1052	816	0.420096	0.804993	0.473170	False	Platynereis- Image
0.7	iou	347	469	583	0.574755	0.445817	0.335239	0.502141	1052	816	0.370229	0.830449	0.417003	False	Platynereis- Image
0.8	iou	500	316	736	0.387255	0.300380	0.203608	0.338330	1052	816	0.260545	0.867385	0.293462	False	Platynereis- Image
0.9	iou	741	75	977	0.091912	0.071293	0.041829	0.080300	1052	816	0.065537	0.919260	0.073816	False	Platynereis- Image

plot_stat_comparison(taus=taus, stats_list=model_stats+model_stats_images_only, model_names=df_all.Dataset.unique(), metric="IoU")

../_images/ae2158a4f38f78f502a65e0ec9aaa4ade69ec4603e747ce09d274dfbab09a407.png