renaming dP_smooth to flow3D_smooth so that it is more clear what it does

Author: carsen-stringer

cellpose/__main__.py

@@ -219,7 +219,7 @@ def main():
                     interp=(not args.no_interp), normalize=normalize,
                     channel_axis=args.channel_axis, z_axis=args.z_axis,
                     anisotropy=args.anisotropy, niter=args.niter,
-                    dP_smooth=args.dP_smooth)
+                    flow3D_smooth=args.flow3D_smooth)
                 masks, flows = out[:2]
                 if len(out) > 3 and restore_type is None:
                     diams = out[-1]

cellpose/cli.py

@@ -115,7 +115,7 @@ def get_arg_parser():
         "--min_size", required=False, default=15, type=int,
         help="minimum number of pixels per mask, can turn off with -1")
     algorithm_args.add_argument(
-        "--dP_smooth", required=False, default=0, type=float,
+        "--flow3D_smooth", required=False, default=0, type=float,
         help="stddev of gaussian for smoothing of dP for dynamics in 3D, default of 0 means no smoothing")
 
     algorithm_args.add_argument(

cellpose/denoise.py

@@ -28,6 +28,8 @@
         if ctype != "cyto3":
             for ltype in ["per", "seg", "rec"]:
                 MODEL_NAMES.append(f"{ntype}_{ltype}_{ctype}")
+    if ctype != "cyto3":
+        MODEL_NAMES.append(f"aniso_{ctype}")
 
 criterion = nn.MSELoss(reduction="mean")
 criterion2 = nn.BCEWithLogitsLoss(reduction="mean")
@@ -505,7 +507,7 @@ def eval(self, x, batch_size=8, channels=None, channel_axis=None, z_axis=None,
              normalize=True, rescale=None, diameter=None, tile_overlap=0.1,
              augment=False, resample=True, invert=False, flow_threshold=0.4,
              cellprob_threshold=0.0, do_3D=False, anisotropy=None, stitch_threshold=0.0,
-             min_size=15, niter=None, interp=True, bsize=224, dP_smooth=0):
+             min_size=15, niter=None, interp=True, bsize=224, flow3D_smooth=0):
         """
         Restore array or list of images using the image restoration model, and then segment.
 
@@ -547,7 +549,7 @@ def eval(self, x, batch_size=8, channels=None, channel_axis=None, z_axis=None,
             anisotropy (float, optional): for 3D segmentation, optional rescaling factor (e.g. set to 2.0 if Z is sampled half as dense as X or Y). Defaults to None.
             stitch_threshold (float, optional): if stitch_threshold>0.0 and not do_3D, masks are stitched in 3D to return volume segmentation. Defaults to 0.0.
             min_size (int, optional): all ROIs below this size, in pixels, will be discarded. Defaults to 15.
-            dP_smooth (int, optional): if do_3D and dP_smooth>0, smooth flows with gaussian filter of this stddev. Defaults to 0.
+            flow3D_smooth (int, optional): if do_3D and flow3D_smooth>0, smooth flows with gaussian filter of this stddev. Defaults to 0.
             niter (int, optional): number of iterations for dynamics computation. if None, it is set proportional to the diameter. Defaults to None.
             interp (bool, optional): interpolate during 2D dynamics (not available in 3D) . Defaults to True.
             

cellpose/gui/gui.py

@@ -301,7 +301,7 @@ def __init__(self, image=None, logger=None):
 
         self.load_3D = False
         self.stitch_threshold = 0.
-        self.dP_smooth = 0.
+        self.flow3D_smooth = 0.
         self.anisotropy = 1.
         self.min_size = 15
         self.resample = True
@@ -2420,8 +2420,8 @@ def compute_segmentation(self, custom=False, model_name=None, load_model=True):
                 self.stitch_threshold, float) else self.stitch_threshold
             anisotropy = float(self.anisotropy.text()) if not isinstance(
                 self.anisotropy, float) else self.anisotropy
-            dP_smooth = float(self.dP_smooth.text()) if not isinstance(
-                self.dP_smooth, float) else self.dP_smooth
+            flow3D_smooth = float(self.flow3D_smooth.text()) if not isinstance(
+                self.flow3D_smooth, float) else self.flow3D_smooth
             min_size = int(self.min_size.text()) if not isinstance(
                 self.min_size, int) else self.min_size
             resample = self.resample.isChecked() if not isinstance(
@@ -2446,7 +2446,7 @@ def compute_segmentation(self, custom=False, model_name=None, load_model=True):
                     cellprob_threshold=cellprob_threshold,
                     flow_threshold=flow_threshold, do_3D=do_3D, niter=niter,
                     normalize=normalize_params, stitch_threshold=stitch_threshold,
-                    anisotropy=anisotropy, resample=resample, dP_smooth=dP_smooth,
+                    anisotropy=anisotropy, resample=resample, flow3D_smooth=flow3D_smooth,
                     min_size=min_size,
                     progress=self.progress, z_axis=0 if self.NZ > 1 else None)[:2]
             except Exception as e:

cellpose/gui/gui3d.py

@@ -165,20 +165,20 @@ def __init__(self, image=None, logger=None):
         )
         self.segBoxG.addWidget(self.stitch_threshold, b, 4, 1, 1)
 
-        label = QLabel("dP_smooth:")
+        label = QLabel("flow3D_smooth:")
         label.setToolTip(
-            "for 3D volumes, smooth flows by a Gaussian with standard deviation dP_smooth (see docs for details)"
+            "for 3D volumes, smooth flows by a Gaussian with standard deviation flow3D_smooth (see docs for details)"
         )
         label.setFont(self.medfont)
         self.segBoxG.addWidget(label, b, 5, 1, 3)
-        self.dP_smooth = QLineEdit()
-        self.dP_smooth.setText("0.0")
-        self.dP_smooth.setFixedWidth(30)
-        self.dP_smooth.setFont(self.medfont)
-        self.dP_smooth.setToolTip(
-            "for 3D volumes, smooth flows by a Gaussian with standard deviation dP_smooth (see docs for details)"
+        self.flow3D_smooth = QLineEdit()
+        self.flow3D_smooth.setText("0.0")
+        self.flow3D_smooth.setFixedWidth(30)
+        self.flow3D_smooth.setFont(self.medfont)
+        self.flow3D_smooth.setToolTip(
+            "for 3D volumes, smooth flows by a Gaussian with standard deviation flow3D_smooth (see docs for details)"
         )
-        self.segBoxG.addWidget(self.dP_smooth, b, 8, 1, 1)
+        self.segBoxG.addWidget(self.flow3D_smooth, b, 8, 1, 1)
 
         b+=1
         label = QLabel("anisotropy:")

cellpose/io.py

@@ -608,7 +608,7 @@ def save_rois(masks, file_name, multiprocessing=None):
 
 
 def save_masks(images, masks, flows, file_names, png=True, tif=False, channels=[0, 0],
-               suffix="", save_flows=False, save_outlines=False, dir_above=False,
+               suffix="_cp_masks", save_flows=False, save_outlines=False, dir_above=False,
                in_folders=False, savedir=None, save_txt=False, save_mpl=False):
     """ Save masks + nicely plotted segmentation image to png and/or tiff.
 
@@ -630,7 +630,7 @@ def save_masks(images, masks, flows, file_names, png=True, tif=False, channels=[
         png (bool, optional): Save masks to PNG. Defaults to True.
         tif (bool, optional): Save masks to TIF. Defaults to False.
         channels (list, int, optional): Channels used to run Cellpose. Defaults to [0,0].
-        suffix (str, optional): Add name to saved masks. Defaults to "".
+        suffix (str, optional): Add name to saved masks. Defaults to "_cp_masks".
         save_flows (bool, optional): Save flows output from Cellpose.eval. Defaults to False.
         save_outlines (bool, optional): Save outlines of masks. Defaults to False.
         dir_above (bool, optional): Save masks/flows in directory above. Defaults to False.
@@ -709,8 +709,7 @@ def save_masks(images, masks, flows, file_names, png=True, tif=False, channels=[
     with warnings.catch_warnings():
         warnings.simplefilter("ignore")
         for ext in exts:
-
-            imsave(os.path.join(maskdir, basename + "_cp_masks" + suffix + ext), masks)
+            imsave(os.path.join(maskdir, basename + suffix + ext), masks)
 
     if save_mpl and png and MATPLOTLIB and not min(images.shape) > 3:
         # Make and save original/segmentation/flows image

cellpose/models.py

@@ -381,7 +381,7 @@ def __init__(self, gpu=False, pretrained_model=False, model_type=None,
     def eval(self, x, batch_size=8, resample=True, channels=None, channel_axis=None,
              z_axis=None, normalize=True, invert=False, rescale=None, diameter=None,
              flow_threshold=0.4, cellprob_threshold=0.0, do_3D=False, anisotropy=None,
-             dP_smooth=0, stitch_threshold=0.0, 
+             flow3D_smooth=0, stitch_threshold=0.0, 
              min_size=15, max_size_fraction=0.4, niter=None, 
              augment=False, tile_overlap=0.1, bsize=224, 
              interp=True, compute_masks=True, progress=None):
@@ -420,7 +420,7 @@ def eval(self, x, batch_size=8, resample=True, channels=None, channel_axis=None,
             flow_threshold (float, optional): flow error threshold (all cells with errors below threshold are kept) (not used for 3D). Defaults to 0.4.
             cellprob_threshold (float, optional): all pixels with value above threshold kept for masks, decrease to find more and larger masks. Defaults to 0.0.
             do_3D (bool, optional): set to True to run 3D segmentation on 3D/4D image input. Defaults to False.
-            dP_smooth (int, optional): if do_3D and dP_smooth>0, smooth flows with gaussian filter of this stddev. Defaults to 0.
+            flow3D_smooth (int, optional): if do_3D and flow3D_smooth>0, smooth flows with gaussian filter of this stddev. Defaults to 0.
             anisotropy (float, optional): for 3D segmentation, optional rescaling factor (e.g. set to 2.0 if Z is sampled half as dense as X or Y). Defaults to None.
             stitch_threshold (float, optional): if stitch_threshold>0.0 and not do_3D, masks are stitched in 3D to return volume segmentation. Defaults to 0.0.
             min_size (int, optional): all ROIs below this size, in pixels, will be discarded. Defaults to 15.
@@ -467,7 +467,7 @@ def eval(self, x, batch_size=8, resample=True, channels=None, channel_axis=None,
                     interp=interp, flow_threshold=flow_threshold,
                     cellprob_threshold=cellprob_threshold, compute_masks=compute_masks,
                     min_size=min_size, max_size_fraction=max_size_fraction, 
-                    stitch_threshold=stitch_threshold, dP_smooth=dP_smooth,
+                    stitch_threshold=stitch_threshold, flow3D_smooth=flow3D_smooth,
                     progress=progress, niter=niter)
                 masks.append(maski)
                 flows.append(flowi)
@@ -522,9 +522,9 @@ def eval(self, x, batch_size=8, resample=True, channels=None, channel_axis=None,
                 resample=resample, do_3D=do_3D, anisotropy=anisotropy)
 
             if do_3D:    
-                if dP_smooth > 0:
-                    models_logger.info(f"smoothing flows with sigma={dP_smooth}")
-                    dP = gaussian_filter(dP, (0, dP_smooth, dP_smooth, dP_smooth))
+                if flow3D_smooth > 0:
+                    models_logger.info(f"smoothing flows with sigma={flow3D_smooth}")
+                    dP = gaussian_filter(dP, (0, flow3D_smooth, flow3D_smooth, flow3D_smooth))
                 torch.cuda.empty_cache()
                 gc.collect()
 

docs/do3d.rst

@@ -49,7 +49,7 @@ Specify this segmentation format in the notebook with ``do_3D=True`` or in the C
 (with the CLI it will segment all tiffs in the folder as 3D tiffs if possible).
 
 If you see many cells that are fragmented, you can smooth the flows before the dynamics 
-are run in 3D using the ``dP_smooth`` parameter, which specifies the standard deviation of 
+are run in 3D using the ``flow3D_smooth`` parameter, which specifies the standard deviation of 
 a Gaussian for smoothing the flows. The default is 0.0, which means no smoothing. Alternatively/additionally,
 you may want to train a model on 2D slices from your 3D data to improve the segmentation (see below).