phenomenon · claim-type · cross-domain-transfer · mechanism · claim-relation · replication-status · limitation — grounded in the full text| phenomenon | small, variable-shaped structures that a single-stage network misses, and the cascades built to hand it a cropped view of them - which repeat the same low-level features in every model of the chain— “Methods for existing applications rely heavily on multi-stage, cascaded CNNs when the target organs show large inter-patient variation in terms of shape and size. Cascaded frameworks extract a region of interest (ROI) and make dense predictions on that particular ROI. The application areas include cardiac MRI, cardiac CT, abdominal CT… | |
| claim-type | method— “To address this general problem, we propose a simple and yet effective solution, named attention gates (AGs). By incorporating AGs into standard CNN models, model parameters and intermediate feature maps are expected to be utilised more efficiently while minimising the necessity of cascaded models to solve localisation and classification… | |
| method — evidence | — “Gating parameters are initialised so that attention gates pass through feature vectors at all spatial locations. Moreover, we do not require multiple training stages as in hard-attention based approaches therefore simplifying the training procedure.” | |
| object of study — evidence | — “We propose a novel attention gate (AG) model for medical image analysis that automatically learns to focus on target structures of varying shapes and sizes. Models trained with AGs implicitly learn to suppress irrelevant regions in an input image while highlighting salient features useful for a specific task. This enables us to eliminate the… | |
| research question — evidence | — “Methods for existing applications rely heavily on multi-stage, cascaded CNNs when the target organs show large inter-patient variation in terms of shape and size. Cascaded frameworks extract a region of interest (ROI) and make dense predictions on that particular ROI. The application areas include cardiac MRI, cardiac CT, abdominal CT… | |
| contribution — evidence | — “We take the attention approach proposed by a step further by proposing grid-based gating that allows attention gates to be more specific to local regions. This improves performance compared to gating based on a global feature vector.” | |
| transfer | from natural language processing: additive soft attention, carried from sentence-to-sentence machine translation into a convolutional network over medical images, and re-based on an image grid so the coefficients can be local— “We take the attention approach proposed by a step further by proposing grid-based gating that allows attention gates to be more specific to local regions. This improves performance compared to gating based on a global feature vector.” | transfers[0] |
| mechanism | localisation folded into the predictor a cascade spends a whole network learning where the organ is and then throws that network's features away; a gate computes the same decision from features the predictor already has, so the localisation costs 8% more parameters instead of a second model, and the low-level features are extracted once— “Methods for existing applications rely heavily on multi-stage, cascaded CNNs when the target organs show large inter-patient variation in terms of shape and size. Cascaded frameworks extract a region of interest (ROI) and make dense predictions on that particular ROI. The application areas include cardiac MRI, cardiac CT, abdominal CT… | mechanisms[0] |
| mechanism | a gate initialised to do nothing the gating parameters start so that every feature vector passes at every location - the network begins as its own baseline and has to learn to suppress, which is why no multi-stage training schedule is needed and why any difference is attributable to what the gate learned rather than to where it started— “Gating parameters are initialised so that attention gates pass through feature vectors at all spatial locations. Moreover, we do not require multiple training stages as in hard-attention based approaches therefore simplifying the training procedure.” | mechanisms[1] |
| mechanism | coarse context deciding fine attention the gating signal is taken from a coarser grid, where the receptive field is large enough to know where the organ is, and applied to a finer map, where the resolution is enough to draw its boundary - which is the same division of labour a cascade makes between its two networks, done inside one— “We commonly observe that AGs initially have a uniform distribution and pass features at all spatial locations. This is gradually updated and localised towards the targeted organ boundaries. Additionally, at coarser scales AGs provide a rough outline of organs which are gradually refined at finer resolutions. Moreover, by training multiple AGs… | mechanisms[2] |
| mechanism | a normalisation choice that trades sparsity against gradients softmax over the attention map produces sparse, intensity-sensitive attention, which is bad in ultrasound; sigmoid only rescales and lets more information through, and pays for it with gradient saturation. Neither is free and the paper reports both costs— “We empirically observed and reported that using soft-max as the activation function tends to generate a map that is sparsely activated and is overly sensitive to local intensity changes. The latter is problematic as in ultrasound imaging, image quality is often low. In the classification setting, We found that dividing the activations by the… | mechanisms[3] |
| critiques | cascaded multi-stage CNN frameworks, which reach 81-82% DSC on the TCIA pancreas benchmark but re-extract the same low-level features in every model of the chain— “Methods for existing applications rely heavily on multi-stage, cascaded CNNs when the target organs show large inter-patient variation in terms of shape and size. Cascaded frameworks extract a region of interest (ROI) and make dense predictions on that particular ROI. The application areas include cardiac MRI, cardiac CT, abdominal CT… | claim_relations[0] |
| critiques | hard-attention approaches, which crop, are non-differentiable, and need reinforcement learning for parameter updates— “Trainable attention, on the other hand, is enforced by design and categorised as hard- and soft-attention. Hard attention, e.g. iterative region proposal and cropping, is often non-differentiable and relies on reinforcement learning for parameter updates, which makes model training more difficult. used recursive hard-attention to detect… | claim_relations[1] |
| extends | soft attention as used for adaptive feature pooling in natural image classification and captioning, generalised here to grid-based gating so coefficients can be specific to local regions— “We take the attention approach proposed by a step further by proposing grid-based gating that allows attention gates to be more specific to local regions. This improves performance compared to gating based on a global feature vector.” | claim_relations[2] |
| qualifies | its own state-of-the-art comparison, which the authors say is not sufficient to draw a final conclusion because the 3D and 2D models are given different information— “Lastly, we note that the presented quantitative comparisons between the Attention 3D-Unet and state-of-the-art 2D cascaded models might not be sufficient enough to draw a final conclusion, as the proposed approach takes advantage of rich contextual… | claim_relations[3] |
| qualifies | the interpretability claim for attention maps - they outline what the model used, not the object, and the paper says so in the same paragraph that offers them for localisation Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[4] |
| sense | an attention map - AS THE PAPER USES IT: a soft region proposal that suppresses background and lets the network localise without a second model Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[0] |
| sense | an attention map - AS THE PAPER SAYS IT IS NOT: the object. It outlines the DISCRIMINANT REGION and does not necessarily coincide with the whole structure Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[1] |
| sense | the gain from attention - AGAINST THE SAME-SIZE BASELINE: 2-3 DSC points for 8% extra parameters Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[2] |
| sense | the gain from attention - AGAINST A BASELINE GROWN TO MATCH: still significant at p = .007, which is what makes it a claim about the gate rather than about size Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[3] |
| sense | the state-of-the-art comparison - AS THE TABLE READS: comparable to 2D cascaded models and better than single-model frameworks Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[4] |
| sense | the state-of-the-art comparison - AS THE AUTHORS QUALIFY IT: not sufficient to draw a final conclusion, because their 3D model trades resolution for context and the 2D models trade context for resolution Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | senses[5] |
| control | higher-capacity U-Net models trained and compared against the smaller gated model, rather than only against the same-size baseline — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. — rules out the extra parameters as the explanation of the gain - a module that adds 8% capacity and improves DSC by 2-3% has to be shown to beat 8% more capacity spread uniformly, and it does at p = .007 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[0] |
| control | the same two models retrained at a smaller training size (30 images instead of 120) — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. — rules out the improvement being an artefact of one data budget - if attention only helps when data is plentiful it is not solving the localisation problem it claims to Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[1] |
| control | models trained on CT-150 applied directly to the public CT-82 benchmark with no fine-tuning, before any adaptation is done — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. — rules out the result being tied to one institution's data - the transfer is measured before the models are allowed to see the new distribution Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[2] |
| control | both models also trained from scratch on 61 CT-82 images, in addition to the fine-tuned and directly-applied conditions — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. — rules out pre-training on the private dataset as a hidden advantage in the public benchmark comparison Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[3] |
| control | 5-fold cross-validation on CT-82 with the Attention U-Net, reported with its standard deviation — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. — rules out a single favourable train/test split - the spread of 6.23 DSC is the size of the effect being claimed, and reporting it makes that visible Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[4] |
| control | conditional random field post-processing deliberately NOT applied, and said so, although the comparison methods use it — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. — rules out post-processing absorbing or masking the gate's contribution - the measured difference is the module's in an isolated setting, at the cost of a lower absolute number Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[5] |
| control | ultrasound frames cropped to the central 208 x 272 so the network cannot read the annotations the scanner prints around the image — held — rules out the network reading the annotations the scanner prints around the image instead of the anatomy - a leak that would look like accuracy Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[6] |
| control | gating parameters initialised so that the gates pass every feature vector at every location, making the network start as its own baseline — held — rules out that the attention gates help only because they change the initialisation - starting open makes the network its own baseline at step zero Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[7] |
| control | class imbalance handled by weighted sampling for classification and a Dice loss for segmentation, in both the attention and the baseline networks — held — rules out the dominant background class carrying the score, and any difference between the two networks being a difference in how each one met the imbalance Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[8] |
| weight | cascaded multi-stage CNN frameworks, which reach 81-82% DSC on the TCIA pancreas benchmark but re-extract the same low-level features in every model of the chain — 0.95 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[0] |
| weight | hard-attention approaches, which crop, are non-differentiable, and need reinforcement learning for parameter updates — 0.95 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[1] |
| weight | soft attention as used for adaptive feature pooling in natural image classification and captioning, generalised here to grid-based gating so coefficients can be specific to local regions — 0.95 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[2] |
| weight | its own state-of-the-art comparison, which the authors say is not sufficient to draw a final conclusion because the 3D and 2D models are given different information — 0.95 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[3] |
| weight | the interpretability claim for attention maps - they outline what the model used, not the object, and the paper says so in the same paragraph that offers them for localisation — 0.9 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | claim_relations[4] |
| audit | attention gates —distinct-from→ cascaded — Audit 2026-09-28: DOUBTFUL - read against its span and kept, but attention gates remove the need for an external localisation module; distinct-from 'cascaded' is terse. Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concept_edges[0] |
| differs in (stated by the tagger, no quotation) | attention gates —distinct-from→ cascaded — whether localisation is a separate model with its own parameters or a module inside the predictor | concept_edges[0] |
| relation | attention gates —distinct-from→ cascaded — the gates suppress background progressively inside one network instead of cropping a region of interest between two Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concept_edges[0] |
| weight | attention gates —distinct-from→ cascaded — 0.95 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concept_edges[0] |
| differs in (stated by the tagger, no quotation) | soft attention —distinct-from→ Hard attention — whether the attention step can be differentiated through | concept_edges[1] |
| relation | soft attention —distinct-from→ Hard attention — hard attention crops and is non-differentiable, needing reinforcement learning; soft attention is probabilistic and trains by ordinary backpropagation Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concept_edges[1] |
| weight | soft attention —distinct-from→ Hard attention — 0.95 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concept_edges[1] |
| audit | attention gates —distinct-from→ trainable parameters — Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span writes 'AGs' and says they contribute more than simply increasing capacity; distinct-from is loose. Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concept_edges[3] |
| differs in (stated by the tagger, no quotation) | attention gates —distinct-from→ trainable parameters — whether the extra parameters are spread across all layers or spent on localisation | concept_edges[3] |
| relation | attention gates —distinct-from→ trainable parameters — a U-Net grown uniformly to the same size does not reach the gated model, so the gain is not capacity Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concept_edges[3] |
| weight | attention gates —distinct-from→ trainable parameters — 0.95 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concept_edges[3] |
| outcome | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[0] |
| outcome | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[1] |
| outcome | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[2] |
| outcome | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[3] |
| outcome | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[4] |
| outcome | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[5] |
| outcome (stated by the tagger, no quotation) | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[6] |
| outcome (stated by the tagger, no quotation) | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[7] |
| outcome (stated by the tagger, no quotation) | control — held Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | controls[8] |
| limitation | the attention map outlines the discriminant region and does not necessarily coincide with the entire object, so it localises without delineating Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[0] |
| limitation | the comparison against 2D cascaded state-of-the-art models is called insufficient to draw a final conclusion, because the 3D model has context and the 2D models have resolution Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[1] |
| limitation | softmax normalisation gives sparse, intensity-sensitive maps and sigmoid gives gradient saturation - the normalisation choice has no free option Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[2] |
| limitation | training the attention mechanism is more complex than the baseline, and the most stable route was per-scale training followed by fine-tuning Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[3] |
| limitation | residual connections around the gate gave no significant improvement in the segmentation experiments Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | limitations[4] |
finding · quantity · variable · definition · key-term — grounded in the full text| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[0] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[1] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[2] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[3] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[4] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[5] |
| finding | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.findings[6] |
| quantity | extra capacity added to the standard U-Net by the attention gates, for a 2-3% DSC improvement Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.quantities[0] |
| quantity | mean pancreas DSC over 5-fold cross-validation on CT-82, with a standard deviation of 6.23 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.quantities[1] |
| quantity | the bottom of the atlas-method DSC range on the TCIA pancreas benchmark Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.quantities[2] |
| variable | recall Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.variables[0] |
| variable | trainable parameters Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.variables[1] |
| definition | soft attention Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.definitions[0] |
| definition | discriminant region Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.definitions[1] |
| concept | attention gates — not defined in this paper's own words; modules inside a single CNN that learn per-location coefficients suppressing irrelevant feature responses, replacing an external localisation network Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[0] |
| concept | cascaded — not defined in this paper's own words; a chain of networks where the first extracts a region of interest and the second predicts inside it - accurate, and repeating the same low-level features in every stage Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[1] |
| concept | soft attention — not defined in this paper's own words; probabilistic, end-to-end differentiable attention trained by standard backpropagation - as against hard attention, which crops and needs reinforcement learning Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[2] |
| concept | Hard attention — not defined in this paper's own words; iterative region proposal and cropping; non-differentiable, and the thing this paper's gates are offered in place of Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[3] |
| concept | discriminant region — not defined in this paper's own words; the part of an object the classifier actually uses - what an attention map outlines, and explicitly not the whole object Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[4] |
| concept | recall — not defined in this paper's own words; the share of true foreground pixels recovered - the quantity attention gates move most in segmentation, at p = .005 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[5] |
| concept | trainable parameters — not defined in this paper's own words; model capacity - the rival explanation for any gain from adding a module, controlled here by growing the baseline instead Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[6] |
| concept | DSC — not defined in this paper's own words; Dice similarity coefficient - the segmentation overlap measure this literature is ranked on, from 69.6% for atlas methods to 81-82% for cascades Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[7] |
| concept | conditional random field — not defined in this paper's own words; a post-processing step common in this literature and deliberately NOT used here, so that the measured difference is the gate's Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[8] |
| DSC | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[0] |
| p | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[1] |
When we have changed our mind. Some entries in the table begin with a word in capitals and the reason for it. EVIDENCE CORRECTED — the finding still stands, but we had quoted the wrong sentence from the paper, and the quotation has been replaced with the right one. RETRACTED — the entry itself was wrong; it stays on the page, marked, with the reason, rather than being deleted. WITHDRAWN AND REPLACED — the entry was wrong, and a corrected version of it appears in the entries below. Each mark carries its own reason, in brackets, because the reasons differ from one entry to the next. An entry we have WITHDRAWN is removed from this page and kept in the record, with its reason. A line reading Audit and a date is a later re-reading of the entry against its whole sentence: DOUBTFUL means it was kept, and the reason says what is loose about it. Quotation marks inside the tagger's sentence. Where a tag is the tagger's own sentence, a phrase in quotation marks inside it is borrowed from the paper — the tagger's wording around it is not. The paper's own sentence is the evidence quote for that row; on this page it is withheld with the article, and the article is at its publisher. Nothing inside quotation marks anywhere on this page is ours: if it is quoted, it is the paper's. Whose words a concept carries. A concept entry records a term the paper itself uses, so the term, the sense given for it, and its evidence are the paper's own words, quoted where the paper fixes them. This is the opposite of a sense entry in the tag table, which is the tagger's sentence about how the paper uses a term. Where a concept entry could not be grounded in the paper's words, the entry says so on its own row.
Tag layer © the Meta-Tagging Project, CC BY 4.0. The paper text is © its authors/publisher
and is linked, not reproduced. Excerpts shown are short quotations for identification and verification.
· back to index
· concepts · authors
· עברית
© 2026 Shir Sivroni · ORCID 0009-0000-2597-2824