Python utils.blob.deserialize() Examples

The following are 10 code examples of utils.blob.deserialize(). You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may also want to check out all available functions/classes of the module utils.blob , or try the search function .
Example #1
Source File: generate_proposal_labels.py    From seg_every_thing with Apache License 2.0 6 votes vote down vote up 
def forward(self, inputs, outputs):
        """See modeling.detector.GenerateProposalLabels for inputs/outputs
        documentation.
        """
        # During training we reuse the data loader code. We populate roidb
        # entries on the fly using the rois generated by RPN.
        # im_info: [[im_height, im_width, im_scale], ...]
        rois = inputs[0].data
        roidb = blob_utils.deserialize(inputs[1].data)
        im_info = inputs[2].data
        im_scales = im_info[:, 2]
        output_blob_names = roi_data.fast_rcnn.get_fast_rcnn_blob_names()
        # For historical consistency with the original Faster R-CNN
        # implementation we are *not* filtering crowd proposals.
        # This choice should be investigated in the future (it likely does
        # not matter).
        json_dataset.add_proposals(roidb, rois, im_scales, crowd_thresh=0)
        blobs = {k: [] for k in output_blob_names}
        roi_data.fast_rcnn.add_fast_rcnn_blobs(blobs, im_scales, roidb)
        for i, k in enumerate(output_blob_names):
            blob_utils.py_op_copy_blob(blobs[k], outputs[i])
Example #2
Source File: generate_proposal_labels.py    From masktextspotter.caffe2 with Apache License 2.0 6 votes vote down vote up 
def forward(self, inputs, outputs):
        """See modeling.detector.GenerateProposalLabels for inputs/outputs
        documentation.
        """
        # During training we reuse the data loader code. We populate roidb
        # entries on the fly using the rois generated by RPN.
        # im_info: [[im_height, im_width, im_scale], ...]
        rois = inputs[0].data
        roidb = blob_utils.deserialize(inputs[1].data)
        im_info = inputs[2].data
        im_scales = im_info[:, 2]
        output_blob_names = roi_data.fast_rcnn.get_fast_rcnn_blob_names()
        # For historical consistency with the original Faster R-CNN
        # implementation we are *not* filtering crowd proposals.
        # This choice should be investigated in the future (it likely does
        # not matter).
        json_dataset.add_proposals(roidb, rois, im_scales, crowd_thresh=0)
        blobs = {k: [] for k in output_blob_names}
        roi_data.fast_rcnn.add_fast_rcnn_blobs(blobs, im_scales, roidb)
        for i, k in enumerate(output_blob_names):
            blob_utils.py_op_copy_blob(blobs[k], outputs[i])
Example #3
Source File: generate_proposal_labels.py    From NucleiDetectron with Apache License 2.0 6 votes vote down vote up 
def forward(self, inputs, outputs):
        """See modeling.detector.GenerateProposalLabels for inputs/outputs
        documentation.
        """
        # During training we reuse the data loader code. We populate roidb
        # entries on the fly using the rois generated by RPN.
        # im_info: [[im_height, im_width, im_scale], ...]
        rois = inputs[0].data
        roidb = blob_utils.deserialize(inputs[1].data)
        im_info = inputs[2].data
        im_scales = im_info[:, 2]
        output_blob_names = roi_data.fast_rcnn.get_fast_rcnn_blob_names()
        # For historical consistency with the original Faster R-CNN
        # implementation we are *not* filtering crowd proposals.
        # This choice should be investigated in the future (it likely does
        # not matter).
        json_dataset.add_proposals(roidb, rois, im_scales, crowd_thresh=0)
        blobs = {k: [] for k in output_blob_names}
        roi_data.fast_rcnn.add_fast_rcnn_blobs(blobs, im_scales, roidb)
        for i, k in enumerate(output_blob_names):
            blob_utils.py_op_copy_blob(blobs[k], outputs[i])
Example #4
Source File: collect_and_distribute_fpn_rpn_proposals.py    From DetectAndTrack with Apache License 2.0 6 votes vote down vote up 
def forward(self, inputs, outputs):
        # inputs is
        # [rpn_rois_fpn2, ..., rpn_rois_fpn6,
        #  rpn_roi_probs_fpn2, ..., rpn_roi_probs_fpn6]
        # If training with Faster R-CNN, then inputs will additionally include
        #  + [roidb, im_info]
        rois = collect(inputs, self._train)
        if self._train:
            # During training we reuse the data loader code. We populate roidb
            # entries on the fly using the rois generated by RPN.
            # im_info: [[im_height, im_width, im_scale], ...]
            im_info = inputs[-1].data
            im_scales = im_info[:, 2]
            roidb = blob_utils.deserialize(inputs[-2].data)
            output_blob_names = roi_data.fast_rcnn.get_fast_rcnn_blob_names()
            json_dataset.add_proposals(roidb, rois, im_scales)
            blobs = {k: [] for k in output_blob_names}
            roi_data.fast_rcnn.add_fast_rcnn_blobs(blobs, im_scales, roidb)
            for i, k in enumerate(output_blob_names):
                blob_utils.py_op_copy_blob(blobs[k], outputs[i])
        else:
            # For inference we have a special code path that avoids some data
            # loader overhead
            distribute(rois, None, outputs, self._train)
Example #5
Source File: collect_and_distribute_fpn_rpn_proposals.py    From seg_every_thing with Apache License 2.0 5 votes vote down vote up 
def forward(self, inputs, outputs):
        """See modeling.detector.CollectAndDistributeFpnRpnProposals for
        inputs/outputs documentation.
        """
        # inputs is
        # [rpn_rois_fpn2, ..., rpn_rois_fpn6,
        #  rpn_roi_probs_fpn2, ..., rpn_roi_probs_fpn6]
        # If training with Faster R-CNN, then inputs will additionally include
        #  + [roidb, im_info]
        rois = collect(inputs, self._train)
        if self._train:
            # During training we reuse the data loader code. We populate roidb
            # entries on the fly using the rois generated by RPN.
            # im_info: [[im_height, im_width, im_scale], ...]
            im_info = inputs[-1].data
            im_scales = im_info[:, 2]
            roidb = blob_utils.deserialize(inputs[-2].data)
            # For historical consistency with the original Faster R-CNN
            # implementation we are *not* filtering crowd proposals.
            # This choice should be investigated in the future (it likely does
            # not matter).
            json_dataset.add_proposals(roidb, rois, im_scales, crowd_thresh=0)
            # Compute training labels for the RPN proposals; also handles
            # distributing the proposals over FPN levels
            output_blob_names = roi_data.fast_rcnn.get_fast_rcnn_blob_names()
            blobs = {k: [] for k in output_blob_names}
            roi_data.fast_rcnn.add_fast_rcnn_blobs(blobs, im_scales, roidb)
            for i, k in enumerate(output_blob_names):
                blob_utils.py_op_copy_blob(blobs[k], outputs[i])
        else:
            # For inference we have a special code path that avoids some data
            # loader overhead
            distribute(rois, None, outputs, self._train)
Example #6
Source File: collect_and_distribute_fpn_rpn_proposals.py    From masktextspotter.caffe2 with Apache License 2.0 5 votes vote down vote up 
def forward(self, inputs, outputs):
        """See modeling.detector.CollectAndDistributeFpnRpnProposals for
        inputs/outputs documentation.
        """
        # inputs is
        # [rpn_rois_fpn2, ..., rpn_rois_fpn6,
        #  rpn_roi_probs_fpn2, ..., rpn_roi_probs_fpn6]
        # If training with Faster R-CNN, then inputs will additionally include
        #  + [roidb, im_info]
        rois = collect(inputs, self._train)
        if self._train:
            # During training we reuse the data loader code. We populate roidb
            # entries on the fly using the rois generated by RPN.
            # im_info: [[im_height, im_width, im_scale], ...]
            im_info = inputs[-1].data
            im_scales = im_info[:, 2]
            roidb = blob_utils.deserialize(inputs[-2].data)
            # For historical consistency with the original Faster R-CNN
            # implementation we are *not* filtering crowd proposals.
            # This choice should be investigated in the future (it likely does
            # not matter).
            json_dataset.add_proposals(roidb, rois, im_scales, crowd_thresh=0)
            # Compute training labels for the RPN proposals; also handles
            # distributing the proposals over FPN levels
            output_blob_names = roi_data.fast_rcnn.get_fast_rcnn_blob_names()
            blobs = {k: [] for k in output_blob_names}
            roi_data.fast_rcnn.add_fast_rcnn_blobs(blobs, im_scales, roidb)
            for i, k in enumerate(output_blob_names):
                blob_utils.py_op_copy_blob(blobs[k], outputs[i])
        else:
            # For inference we have a special code path that avoids some data
            # loader overhead
            distribute(rois, None, outputs, self._train)
Example #7
Source File: collect_and_distribute_fpn_rpn_proposals_rec.py    From masktextspotter.caffe2 with Apache License 2.0 5 votes vote down vote up 
def forward(self, inputs, outputs):
        """See modeling.detector.CollectAndDistributeFpnRpnProposalsRec for
        inputs/outputs documentation.
        """
        # inputs is
        # [rpn_rois_fpn2, ..., rpn_rois_fpn6,
        #  rpn_roi_probs_fpn2, ..., rpn_roi_probs_fpn6]
        # If training with Faster R-CNN, then inputs will additionally include
        #  + [roidb, im_info]
        rois = collect(inputs, self._train)
        if self._train:
            # During training we reuse the data loader code. We populate roidb
            # entries on the fly using the rois generated by RPN.
            # im_info: [[im_height, im_width, im_scale], ...]
            im_info = inputs[-1].data
            im_scales = im_info[:, 2]
            roidb = blob_utils.deserialize(inputs[-2].data)
            # For historical consistency with the original Faster R-CNN
            # implementation we are *not* filtering crowd proposals.
            # This choice should be investigated in the future (it likely does
            # not matter).
            json_dataset.add_proposals(roidb, rois, im_scales, crowd_thresh=0)
            # Compute training labels for the RPN proposals; also handles
            # distributing the proposals over FPN levels
            output_blob_names = roi_data.fast_rcnn.get_fast_rcnn_blob_names()
            blobs = {k: [] for k in output_blob_names}
            roi_data.fast_rcnn.add_fast_rcnn_blobs_rec(blobs, im_scales, roidb)
            for i, k in enumerate(output_blob_names):
                blob_utils.py_op_copy_blob(blobs[k], outputs[i])
        else:
            # For inference we have a special code path that avoids some data
            # loader overhead
            distribute(rois, None, outputs, self._train)
Example #8
Source File: loader.py    From NucleiDetectron with Apache License 2.0 5 votes vote down vote up 
def save_im_masks(self, blobs):
        import os, uuid
        from datasets.dataset_catalog import _DATA_DIR
        import utils.blob as blob_utils

        channel_swap = (0, 2, 3, 1)
        data = blobs['data'].copy()

        im = data.transpose(channel_swap)[0]
        im = self.rescale_0_1(im)

        roidb_temp = blob_utils.deserialize(blobs['roidb'])[0]

        im_name = str(self._counter) + '_' + os.path.splitext(os.path.basename(roidb_temp['image']))[0]

        with self._lock:
            self._counter += 1

        out_dir = os.path.join(_DATA_DIR, 'vis', roidb_temp['nuclei_class'])
        im_name += '_' + str(uuid.uuid4().get_hex().upper()[0:6])

        try:
            os.makedirs(out_dir)
        except:
            pass

        aug_rles = roidb_temp['segms']

        boxes = roidb_temp['boxes']
        boxes = np.append(boxes, np.ones((len(boxes), 2)), 1)
        im_scale = blobs['im_info'][0, 2]

        from utils.vis import vis_one_image
        vis_one_image(im, im_name, out_dir, boxes, segms=aug_rles, keypoints=None, thresh=0.7,
                      box_alpha=0.8, show_class=False, scale=im_scale)
Example #9
Source File: collect_and_distribute_fpn_rpn_proposals.py    From NucleiDetectron with Apache License 2.0 5 votes vote down vote up 
def forward(self, inputs, outputs):
        """See modeling.detector.CollectAndDistributeFpnRpnProposals for
        inputs/outputs documentation.
        """
        # inputs is
        # [rpn_rois_fpn2, ..., rpn_rois_fpn6,
        #  rpn_roi_probs_fpn2, ..., rpn_roi_probs_fpn6]
        # If training with Faster R-CNN, then inputs will additionally include
        #  + [roidb, im_info]
        rois = collect(inputs, self._train)
        if self._train:
            # During training we reuse the data loader code. We populate roidb
            # entries on the fly using the rois generated by RPN.
            # im_info: [[im_height, im_width, im_scale], ...]
            im_info = inputs[-1].data
            im_scales = im_info[:, 2]
            roidb = blob_utils.deserialize(inputs[-2].data)
            # For historical consistency with the original Faster R-CNN
            # implementation we are *not* filtering crowd proposals.
            # This choice should be investigated in the future (it likely does
            # not matter).
            json_dataset.add_proposals(roidb, rois, im_scales, crowd_thresh=0)
            # Compute training labels for the RPN proposals; also handles
            # distributing the proposals over FPN levels
            output_blob_names = roi_data.fast_rcnn.get_fast_rcnn_blob_names()
            blobs = {k: [] for k in output_blob_names}
            roi_data.fast_rcnn.add_fast_rcnn_blobs(blobs, im_scales, roidb)
            for i, k in enumerate(output_blob_names):
                blob_utils.py_op_copy_blob(blobs[k], outputs[i])
        else:
            # For inference we have a special code path that avoids some data
            # loader overhead
            distribute(rois, None, outputs, self._train)
Example #10
Source File: generate_proposal_labels.py    From DetectAndTrack with Apache License 2.0 5 votes vote down vote up 
def forward(self, inputs, outputs):
        # During training we reuse the data loader code. We populate roidb
        # entries on the fly using the rois generated by RPN.
        # im_info: [[im_height, im_width, im_scale], ...]
        rois = inputs[0].data
        roidb = blob_utils.deserialize(inputs[1].data)
        im_info = inputs[2].data
        im_scales = im_info[:, 2]
        output_blob_names = roi_data.fast_rcnn.get_fast_rcnn_blob_names()
        json_dataset.add_proposals(roidb, rois, im_scales)
        blobs = {k: [] for k in output_blob_names}
        roi_data.fast_rcnn.add_fast_rcnn_blobs(blobs, im_scales, roidb)
        for i, k in enumerate(output_blob_names):
            blob_utils.py_op_copy_blob(blobs[k], outputs[i])