{"id":"31e8a170-c7e7-4400-9435-553d4a35c028","arxiv_id":"2607.06007","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.5,"correctness_risk":"unknown","formal_verification":"none","parameter_count":2,"one_line_summary":"OBBSeg segments irregular medical lesions from oriented bounding-box labels via a Mask-to-OBB loss and prompt modules, claiming near fully-supervised accuracy across 13 datasets and 5 modalities.","lead":"OBBSeg trains medical lesion segmenters from oriented bounding boxes instead of full pixel masks, using a Mask-to-OBB consistency loss and prompt-driven feature modules. If the reported near fully-supervised accuracy holds, it could cut costly pixel labeling while handling elongated lesions better than axis-aligned weak labels.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.5","headline":"Mask-to-OBB only matches oriented extent, not unique irregular boundaries; near-full-supervision parity therefore likely rides on PAFE/DBFE semantic cues rather than OBB geometry alone.","rationale":"The reader’s weakest_assumption correctly isolates the design premise (OBB + Mask-to-OBB as a tight intermediate without residual rectangular bias or heavy prompt dependence). With the methods description in view, that concern is sharper and more technical: Mask-to-OBB is non-unique for irregular shapes, so the empirical “comparable to full supervision” claim is load-bearing on PAFE/DBFE. No circular derivation or formal error; the issue is attribution of the gains. Moving from UNVERDICTED to CONDITIONAL is appropriate: accept the method’s usefulness if the ablation (or existing paper ablations) shows geometry alone is already near full-sup, otherwise require reframing that OBB is a useful cue but not the primary shape constraint. Ordinary checks (how OBBs were obtained, backbone/protocol parity, statistical significance of “comparable”) remain secondary. No ad-hominem; the argument’s soft spot is the non-uniqueness of the geometric loss for the claimed irregular-boundary recovery.","tokens_in":2027,"tokens_out":656,"duration_ms":40539,"concrete_test":"Train an ablated OBBSeg with PAFE and DBFE removed (or features zeroed), using only Mask-to-OBB against OBB labels on the same 13 datasets/splits and backbone. If mean Dice falls >5 points vs. full OBBSeg and clearly below the fully-supervised baseline band in the main tables, the geometric-intermediate premise fails and the claim must be qualified as OBB+prompt rather than OBB-driven. Optionally report mean IoU(pred, minOBB(pred)) vs. IoU(GT, minOBB(GT)) to quantify residual rectangular fill.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim requires that OBB labels plus Mask-to-OBB form a tight enough intermediate to recover irregular lesion masks at near fully-supervised accuracy. Mask-to-OBB is a differentiable min-OBB of the predicted mask matched to the annotated OBB: it constrains support, scale, and orientation, but is satisfied by any mask (rectangle, ellipse, or highly irregular) whose minOBB equals the annotation. Irregular boundary shape is therefore under-determined by the geometric loss. Recovery must come from network bias and especially the PAFE/DBFE prompt modules. Unless ablations show Mask-to-OBB + OBB alone already approaches the full-supervision band, the paradigm’s premise—that OBB geometry is the key intermediate—does not hold; reported parity may be driven by semantic prompts. Residual rectangular bias is also unchecked if predictions systematically fill OBB interiors more than true irregular GT. This is the single load-bearing soft spot of the strongest claim; ordinary ML risks (backbone parity, OBB derivation from GT masks vs. human draw) are secondary.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The manuscript proposes OBBSeg, a weakly supervised segmentation framework that uses oriented bounding boxes (OBBs) as an intermediate supervision signal for irregular lesion segmentation. It introduces a differentiable Mask-to-OBB geometric consistency loss intended to reduce the rectangular bias of OBB labels, together with two prompt-driven modules (PAFE and DBFE) that enhance foreground representation and suppress background. The central empirical claim is that, across 13 datasets spanning five imaging modalities, OBBSeg outperforms existing weakly supervised methods and reaches performance comparable to fully supervised baselines, while requiring only OBB annotations rather than pixel masks.","tokens_in":2286,"tokens_out":1337,"duration_ms":28330,"significance":"If the near-full-supervision parity claim holds under fair controls, the work would be a practically useful contribution: OBB labels are cheaper than masks yet more informative than axis-aligned boxes for elongated or anisotropic lesions, and a reproducible intermediate-supervision recipe would matter for scalable medical segmentation. The paper also ships public code, which strengthens reproducibility. The significance, however, rests on whether OBB geometry plus Mask-to-OBB is genuinely the tight intermediate that recovers irregular boundaries, or whether reported parity is driven mainly by the semantic prompt modules and backbone capacity.","major_comments":[{"comment":"The load-bearing premise is that OBB labels plus the Mask-to-OBB loss supply enough geometric constraint to recover irregular lesion boundaries at near fully-supervised accuracy. Mask-to-OBB matches the min-OBB of the predicted mask to the annotated OBB and therefore constrains support, scale, and orientation, but is satisfied by any mask (rectangle, ellipse, or irregular) whose min-OBB equals the annotation. Irregular boundary shape is under-determined by this loss alone. The manuscript needs a clear ablation isolating (i) OBB + Mask-to-OBB without PAFE/DBFE, (ii) PAFE/DBFE without Mask-to-OBB, and (iii) the full model, reported against the fully supervised upper bound on the same splits. Without that isolation, the claim that OBB geometry is the key intermediate remains unproven and the near-full-supervision parity may be attributable to semantic prompt cues rather than OBB supervision","section":null},{"comment":"Relatedly, residual rectangular bias is not adequately checked. Because Mask-to-OBB is satisfied by masks that fill the OBB interior, predictions may systematically over-cover relative to true irregular ground-truth masks. The paper should report quantitative bias diagnostics (e.g., predicted-mask area / OBB area vs. GT-mask area / OBB area; boundary irregularity metrics such as perimeter-to-area or fractal-like roughness; and qualitative failure cases on highly concave lesions) so that readers can judge whether the method truly recovers irregular shapes or merely produces OBB-filling blobs that score well on overlap metrics.","section":null},{"comment":"The experimental claim of 'performance comparable to fully supervised approaches' on 13 datasets is central and must be stated with precise numbers, variance, and protocol. Report mean±std (or CI) over multiple seeds, the exact fully supervised backbone and training recipe used as the upper bound, and whether OBB labels were derived automatically from GT masks or drawn independently by annotators. Automatic min-OBB-from-GT is a common but optimistic protocol that understates real annotation noise; if that is what was used, it should be disclosed and, ideally, complemented by a human-drawn OBB subset. Statistical tests or at least paired per-dataset deltas versus the strongest weak baseline and the full-supervision band are needed before the parity claim can be accepted.","section":null},{"comment":"The multi-loss design (Mask-to-OBB weight and related coefficients) and the PAFE/DBFE modules introduce free parameters and architectural capacity that are not free under pure OBB supervision. Sensitivity of the main results to the Mask-to-OBB loss weight, and a controlled comparison against a strong box-supervised baseline that uses the same backbone and comparable prompt capacity without OBB-specific geometry, are required to show that gains come from the OBB intermediate rather than from extra modules or tuning.","section":null}],"minor_comments":[{"comment":"Define PAFE and DBFE on first use in the abstract and introduction (Prompt-Aware / Dual-Branch Feature Enhancement or whatever the expansions are) so readers do not have to reverse-engineer acronyms.","section":null},{"comment":"Clarify the exact differentiable construction of Mask-to-OBB (how min-area oriented rectangle is obtained through the network, any relaxations or approximations, and gradient flow) with an equation and a short derivation or pseudocode; the abstract asserts differentiability but the geometric details matter for reproducibility.","section":null},{"comment":"List the 13 datasets and five modalities explicitly in a table with image counts, lesion types, and train/val/test splits so the multi-modality claim is auditable at a glance.","section":null},{"comment":"When claiming outperformance over 'existing weakly supervised methods,' name the primary competitors (e.g., CAM-based, scribble, axis-aligned box methods) and ensure they are re-implemented or re-evaluated under the same backbone and splits where possible.","section":null},{"comment":"Code is linked; please pin the commit/release used for the reported numbers and include training configs for the main tables to make the GitHub artifact match the paper.","section":null}],"recommendation":"major_revision","confidential_remarks":"I only had reliable access to the abstract plus the stress-test framing in this review cycle; if the full PDF contains the requested ablations (Mask-to-OBB alone vs. PAFE/DBFE alone, rectangular-bias diagnostics, and seed-wise full-supervision gaps), several major comments can be downgraded after verification. The core scientific risk is real even with full text: Mask-to-OBB is a support/orientation constraint, not a unique shape prior, so the paper’s narrative that OBB is a tight intermediate must be earned by ablations rather than asserted. Scope fit for a serious CV/medical imaging venue is fine if those controls land."},"author_rebuttal":{"model":"grok-4.5","summary":"We thank the referee for a careful and constructive review. The four major comments correctly identify where the manuscript’s central claim—that OBB geometry plus Mask-to-OBB is a tight intermediate for irregular lesions—must be isolated more rigorously from prompt capacity, residual rectangular bias, protocol details, and free parameters. We agree with the substance of each point and will revise the experimental section accordingly: new ablations that factor OBB/Mask-to-OBB from PAFE/DBFE, quantitative bias diagnostics, multi-seed statistics with explicit full-supervision protocol and OBB provenance, loss-weight sensitivity, and a same-backbone box-supervised control. We believe these additions will make the contribution clearer without changing the core method.","responses":[{"response":"We agree that Mask-to-OBB alone under-determines boundary irregularity: any mask whose minimum OBB matches the annotation satisfies the geometric term, so shape recovery cannot be attributed to OBB geometry without isolating the prompt modules. The current manuscript does not present the three-way factorization the referee requests, and that omission weakens the load-bearing claim. In revision we will add a controlled ablation on the same splits and backbone, reporting (i) OBB + Mask-to-OBB without PAFE/DBFE, (ii) PAFE/DBFE without Mask-to-OBB (using only standard weak box-style or prompt cues as applicable), and (iii) the full OBBSeg model, each against the identical fully supervised upper bound. We will discuss which component closes how much of the gap to full supervision and qualify the “OBB as intermediate” claim accordingly. If (i) alone remains far from the full-supervision band, we will state that OBB geometry is necessary but not sufficient and that semantic prompts carry substantial weight.","revision_made":"yes","referee_comment":"The load-bearing premise is that OBB labels plus the Mask-to-OBB loss supply enough geometric constraint to recover irregular lesion boundaries at near fully-supervised accuracy. Mask-to-OBB matches the min-OBB of the predicted mask to the annotated OBB and therefore constrains support, scale, and orientation, but is satisfied by any mask (rectangle, ellipse, or irregular) whose min-OBB equals the annotation. Irregular boundary shape is under-determined by this loss alone. The manuscript needs a clear ablation isolating (i) OBB + Mask-to-OBB without PAFE/DBFE, (ii) PAFE/DBFE without Mask-to-OBB, and (iii) the full model, reported against the fully supervised upper bound on the same splits. Without that isolation, the claim that OBB geometry is the key intermediate remains unproven and the near-full-supervision parity may be attributable to semantic prompt cues rather than OBB supervision"},{"response":"This concern is well founded. Overlap metrics (Dice/IoU) can remain high for OBB-filling blobs even when boundaries are systematically smoother or more rectangular than the ground truth. The manuscript currently lacks quantitative checks for residual rectangular bias. In revision we will report, on representative datasets spanning elongated and irregular lesions: (1) predicted-mask area / OBB area versus GT-mask area / OBB area (mean and distribution); (2) boundary irregularity measures such as perimeter-to-area ratio and a simple roughness/compactness index for predictions versus GT; and (3) qualitative failure cases focused on highly concave or multi-lobed lesions, including side-by-side comparison with the annotated OBB and the fully supervised prediction. If systematic over-coverage or under-irregularity remains, we will document it and temper claims that Mask-to-OBB fully removes rectangular bias.","revision_made":"yes","referee_comment":"Relatedly, residual rectangular bias is not adequately checked. Because Mask-to-OBB is satisfied by masks that fill the OBB interior, predictions may systematically over-cover relative to true irregular ground-truth masks. The paper should report quantitative bias diagnostics (e.g., predicted-mask area / OBB area vs. GT-mask area / OBB area; boundary irregularity metrics such as perimeter-to-area or fractal-like roughness; and qualitative failure cases on highly concave lesions) so that readers can judge whether the method truly recovers irregular shapes or merely produces OBB-filling blobs that score well on overlap metrics."},{"response":"We accept that the parity claim must be stated with stricter protocol transparency and statistics. In the revised manuscript we will: (1) report mean±std over multiple random seeds for OBBSeg, the strongest weak baselines, and the fully supervised upper bound on the same splits; (2) specify the exact fully supervised backbone, losses, and training recipe used as the upper bound so that capacity is matched; (3) disclose OBB provenance explicitly—if OBBs were obtained as minimum OBBs from GT masks (the common automatic protocol), we will state that and note that it is optimistic relative to independent human OBB annotation; where feasible we will add a small human-drawn OBB subset or a controlled noise study to bound sensitivity to annotation noise; and (4) provide paired per-dataset deltas (and, where appropriate, simple significance tests) versus the strongest weak baseline and versus the full-supervision band. We will rephrase “comparable to fully supervised” to the precise numerical relationship that the multi-seed results support, rather than a blanket parity claim.","revision_made":"yes","referee_comment":"The experimental claim of 'performance comparable to fully supervised approaches' on 13 datasets is central and must be stated with precise numbers, variance, and protocol. Report mean±std (or CI) over multiple seeds, the exact fully supervised backbone and training recipe used as the upper bound, and whether OBB labels were derived automatically from GT masks or drawn independently by annotators. Automatic min-OBB-from-GT is a common but optimistic protocol that understates real annotation noise; if that is what was used, it should be disclosed and, ideally, complemented by a human-drawn OBB subset. Statistical tests or at least paired per-dataset deltas versus the strongest weak baseline and the full-supervision band are needed before the parity claim can be accepted."},{"response":"We agree that free parameters and extra module capacity must be controlled before attributing gains to the OBB intermediate. In revision we will include: (1) a sensitivity study of the main metrics to the Mask-to-OBB loss weight (and related multi-loss coefficients) over a reasonable range, reporting stability or degradation; and (2) a controlled comparison against a strong axis-aligned (or generic box) supervised baseline that uses the same backbone and comparable prompt-style capacity (PAFE/DBFE or an equivalent prompt pathway) but without OBB-specific geometry or the Mask-to-OBB term. This isolates whether orientation-aware OBB supervision and Mask-to-OBB add value beyond backbone capacity and semantic prompts. Results will be discussed honestly: if the OBB-specific path yields only marginal gains under matched capacity, we will revise the narrative to emphasize the joint system rather than OBB geometry alone.","revision_made":"yes","referee_comment":"The multi-loss design (Mask-to-OBB weight and related coefficients) and the PAFE/DBFE modules introduce free parameters and architectural capacity that are not free under pure OBB supervision. Sensitivity of the main results to the Mask-to-OBB loss weight, and a controlled comparison against a strong box-supervised baseline that uses the same backbone and comparable prompt capacity without OBB-specific geometry, are required to show that gains come from the OBB intermediate rather than from extra modules or tuning."}],"tokens_in":2030,"tokens_out":1679,"duration_ms":34779,"standing_objections":[]},"desk_editor":{"model":"grok-4.5","letter":"The one thing to know: OBBSeg is a coherent weak-supervision pipeline that treats oriented boxes as an intermediate annotation tier for elongated medical lesions, with a differentiable Mask-to-OBB consistency loss plus two prompt modules (PAFE, DBFE). The load-bearing claim is empirical parity with full supervision across 13 datasets and 5 modalities. That claim is only as strong as the ablations that separate geometry from the semantic prompts.\n\nWhat is actually new is the packaging, not a first-principles breakthrough. Oriented boxes are standard in detection; box-to-mask weak supervision is standard too. Combining them for irregular lesions with a min-OBB matching loss and prompt-driven foreground/background modules is solid engineering. They cover multiple modalities, ship code, and frame OBB as cheaper than masks and tighter than axis-aligned boxes. That framing is useful for people who already draw boxes. Credit where due: multi-dataset empirical work with public code is real work.\n\nThe soft spot is the one the stress-test names, and it lands. Mask-to-OBB only forces the predicted mask’s minimum oriented box to match the annotation. That constrains support, scale, and angle; it is satisfied by any shape—rectangle, ellipse, or true irregular mask—whose minOBB equals the label. Irregular boundaries are under-determined by the geometric loss alone. Near-full-supervision parity therefore almost certainly depends on PAFE/DBFE and network bias more than on OBB geometry as the “key intermediate.” Residual rectangular fill bias is also unchecked without boundary-error or qualitative analysis. Secondary issues are ordinary ML ones: multi-loss weights, whether OBBs were derived from GT masks or drawn by humans, backbone fairness. None of these kill the paper; they are what a referee should demand.\n\nThis is for medical segmentation researchers who care about annotation cost and anisotropic lesions. A general CV reading group can skip it; a medical imaging group should look. It deserves a serious referee—multi-dataset methods papers with code are exactly what peer review is for. I would not desk-reject.","headline":"Practical OBB weak-supervision stack for irregular lesions; the near-full-supervision claim likely rides more on PAFE/DBFE than on OBB geometry alone.","tokens_in":2934,"tokens_out":528,"would_cite":false,"duration_ms":25044,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"OBBSeg turns oriented bounding boxes into near full-supervision accuracy for irregular lesion segmentation across 13 medical datasets.","keywords":["oriented bounding boxes","weak supervision","medical image segmentation","lesion segmentation","geometric consistency","Mask-to-OBB loss","prompt-driven modules","irregular lesions"],"falsifier":"On a held-out collection of highly concave or multi-component lesions, measure whether OBBSeg Dice remains well below a fully supervised baseline while residual rectangular artifacts persist inside the predicted masks; a large, systematic gap would falsify the claim that OBB geometry is a tight enough intermediate.","tokens_in":2886,"feed_emoji":"🩻","tokens_out":821,"duration_ms":22858,"temperature":0.7,"pith_summary":"Pixel-level labels remain expensive in medical image segmentation, so weak supervision is attractive but usually too loose for irregular lesions. This paper introduces OBBSeg, an intermediate paradigm that uses oriented bounding boxes to supply both spatial extent and orientation, better matching elongated or anisotropic lesions than ordinary axis-aligned boxes. A differentiable Mask-to-OBB loss forces predicted masks to stay geometrically consistent with those boxes and thereby reduces rectangular bias, while two prompt modules (PAFE and DBFE) sharpen foreground cues and suppress background. Experiments on 13 datasets spanning five imaging modalities show that the approach outperforms prior weakly supervised methods and reaches accuracy comparable to fully supervised training. The result offers a practical route to scale medical segmentation without dense pixel painting.","feed_headline":"Oriented boxes nearly match full labels for lesion masks","feed_subtitle":"A Mask-to-OBB loss plus prompts closes the gap across 13 medical datasets and five modalities","key_machinery":"The Mask-to-OBB loss: a differentiable geometric consistency term that maps the predicted mask back onto the oriented-box region so the network cannot collapse to a rectangular bias. Two complementary prompt modules (PAFE for foreground enhancement and DBFE for background suppression) add semantic guidance that further tightens the mask.","core_discovery":"Oriented bounding boxes, when paired with a differentiable Mask-to-OBB geometric consistency loss and prompt-driven semantic modules, supply enough shape and orientation constraint to recover irregular lesion boundaries at a level comparable to full pixel-level supervision while remaining far cheaper to annotate.","pith_inferences":["If a single OBB is already intermediate enough, similar oriented-box supervision may transfer to non-medical elongated structures such as vessels, cracks or roads.","The residual gap to full supervision, if any, is likely largest for highly concave or multi-part lesions where one OBB under-constrains topology.","An active-learning loop that escalates only ambiguous OBBs to full masks could cut annotation cost still further while preserving accuracy.","Differentiable geometric consistency of this form could be applied to other intermediate annotations such as ellipses or sparse polygons."],"forward_implications":["Clinical pipelines can replace dense pixel painting with cheaper OBB drawing for many irregular-lesion tasks while retaining near-full-supervision accuracy.","Weakly supervised methods that currently rely only on axis-aligned boxes can be upgraded by switching to oriented boxes plus geometric consistency.","The same recipe generalizes across at least five imaging modalities and thirteen datasets without modality-specific redesign.","Annotation budgets can be redirected from exhaustive masks toward rapid OBB labels, lowering the cost of building large medical segmentation datasets."],"fun_headline_variants":["Oriented boxes with Mask-to-OBB loss nearly match full lesion labels","Geometric Mask-to-OBB loss makes oriented boxes rival full supervision","Oriented bounding boxes recover irregular lesions like full pixel labels","OBBSeg: prompts and OBB loss close gap to full medical lesion masks","Intermediate OBB supervision matches full labels for anisotropic lesions"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"That oriented boxes plus the geometric consistency loss already give enough shape constraint to recover irregular lesion boundaries without leftover rectangular bias or heavy dependence on the semantic prompt modules.","fun_headline_variants_meta":{"raw":{"variants":["Oriented boxes with Mask-to-OBB loss nearly match full lesion labels","Geometric Mask-to-OBB loss makes oriented boxes rival full supervision","Oriented bounding boxes recover irregular lesions like full pixel labels","OBBSeg: prompts and OBB loss close gap to full medical lesion masks","Intermediate OBB supervision matches full labels for anisotropic lesions"]},"model":"grok-4.5","cost_usd":0.012148,"raw_usage":{"total_tokens":2564,"prompt_tokens":713,"num_sources_used":0,"completion_tokens":74,"cost_in_usd_ticks":121480000,"prompt_tokens_details":{"text_tokens":713,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":1777,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":713,"tokens_out":74,"duration_ms":19319,"temperature":1.0,"reasoning_tokens":1777,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-08T20:04:47.230480+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"On a held-out collection of highly concave or multi-component lesions, measure whether OBBSeg Dice remains well below a fully supervised baseline while residual rectangular artifacts persist inside the predicted masks; a large, systematic gap would falsify the claim that OBB geometry is a tight enough intermediate.","supporting_citations":[],"review_version":1}