Detection, segmentation, depth and recognition research, plus the vision backbones that keep displacing the last generation.
arXiv:2608. 19973v1 Announce Type: cross Abstract: Recently, open-vocabulary 3D object detection (3D-OVD) has gained increasing attention for its ability to detect unseen objects in 3D scenes.
arXiv:2608. 20305v1 Announce Type: new Abstract: Myocardial scar segmentation from single-stack late gadolinium-enhanced cardiac magnetic resonance (LGE-CMR) imaging has been a longstanding and clinically important challenge, particularly in the presence of low tissue contrast, diffuse, and small scar regions.
Reference-based diffusion stylization requires separating target geometry from transferable appearance. Existing tuning-based methods often rely on aligned content-style-target triplets or auxiliary visual encoders, which increases data cost and can transfer unintended scene structure from the style reference.
The paper introduces HN-CLIP, a new objective for dense-caption retrieval that adapts similarity margins per negative example using the text encoder’s own geometry. By adding a detached caption‑similarity matrix to the negative logits, HN‑CLIP addresses the issue of near‑duplicate captions that cause premature loss saturation in InfoNCE training. Experiments on four benchmarks show that HN‑CLIP outperforms leading methods by 2.4–4.3 R@1, trains 2.4× faster than GOAL and 5.4× faster than StructXLIP, and achieves the best full‑data baseline with only 20% of the training data.
The paper proposes a systematic framework for creating a "Map of Datasets in Engineering Design and Systems Engineering" (EDSE) to address the fragmented and inaccessible nature of existing datasets. It introduces a multi‑dimensional taxonomy that classifies datasets by domain, lifecycle stage, data type, and format, and presents an interactive discovery tool built on a knowledge graph data model. The authors analyze the current data landscape, identify underrepresented areas such as early‑stage design and system architecture, and suggest strategies for curation and sustainability to build a dynamic, community‑driven resource.
The paper presents a method for extracting key information from OCR‑digitized clinical reports, addressing challenges posed by heterogeneous documents and noisy OCR output. It introduces an open key space that is iteratively mined, normalized, clustered, and verified to build a canonical key inventory, and defines key coverage as a metric for inventory completeness. Experiments on reports from over 20 hospitals using a 0.2B BERT model show that performance improves steadily with key coverage, achieving high F1 scores when the top 90 keys are covered and outperforming a fine‑tuned Qwen3‑0.6B baseline.
MedStruct‑S is a benchmark for semi‑structured information extraction from OCR‑derived clinical reports, covering key discovery, key‑conditioned QA, and end‑to‑end key‑value extraction. It contains 3,582 annotated real‑world report pages and evaluates models under unknown keys and OCR noise. Experiments show encoder‑only models excel at non‑null key‑conditioned QA, while fine‑tuned decoder‑only models achieve the strongest overall performance across model sizes.
The study investigates how few expert-annotated cases are needed to fine‑tune MedSAM3 for abdominal organ segmentation using Low‑Rank Adaptation (LoRA). With only 10 annotated CT or MRI cases, the LoRA‑adapted models achieve performance comparable to specialist systems that require orders of magnitude more data, including reliable gallbladder segmentation and near‑state‑of‑the‑art results for liver, kidneys, and spleen. The approach also generalizes to cardiac segmentation on the Whole Heart dataset, and training takes only 3–5 hours per organ on a single GPU, roughly twice as fast as nnU-Net.
arXiv:2608. 18339v1 Announce Type: cross Abstract: Vision-language models (VLMs) have demonstrated remarkable zero-shot capabilities yet remain sensitive to real-world distribution shifts during inference.
The paper introduces DistScan, a backdoor detection framework for object detection models that identifies malicious behavior by detecting shifts in the pre‑NMS prediction class distribution relative to training class frequencies. DistScan operates on clean validation data, requiring no access to model weights, trigger knowledge, or additional training, and it aggregates intermediate predictions to flag backdoored models. Experiments on MS‑COCO and PASCAL VOC across two architectures and three scene‑level attack scenarios show that DistScan outperforms existing methods, improving average detection accuracy by 27.32 percentage points over the best baseline.
OmniHandwritingOCR is a diagnostic benchmark designed to evaluate multimodal large language models (MLLMs) and OCR systems on handwritten text and mathematical expression recognition. It comprises 77.57K labeled images across six subtasks and twelve subsets, including a difficulty‑stratified multi‑line formula corpus that tests robustness to increasing structural complexity. The benchmark reveals that current systems perform poorly on complex multi‑line formulas, exhibit variable rankings across languages and formula settings, and sometimes hallucinate corrections that are not visually supported.
OptiModNet is a lightweight UNet‑Transformer hybrid designed for optic disc and cup segmentation. It incorporates grouped‑query and channel attention across multiple stages, along with an Aggregated Pyramid Loss to improve gradient flow and structural consistency. Evaluated on the REFUGE2 dataset, it surpasses existing methods by over 2.5 % while using only 3.73 GFLOPs and 1.93 M parameters.
The paper presents a distribution‑free risk‑control framework that provides organ‑specific recall guarantees for frozen multi‑organ CT segmentation models. It calibrates per‑organ thresholds for an AMOS‑trained nnU‑Net, audits its transfer to RAOS, and estimates local re‑certification costs using case‑level voxel false‑negative rates. The study compares Risk‑Controlling Prediction Sets (RCPS) and Conformal Risk Control (CRC), noting that RCPS offers high‑probability control of population‑mean risk while CRC provides weaker expectation control, and evaluates the effectiveness of the Waudby–Smith–Ramdas betting bound versus Hoeffding–Bentkus bounds for re‑certification. "whyItMatters":"The work demonstrates how to maintain organ‑level recall guarantees when deploying segmentation models across different clinical domains, highlighting the trade‑offs between threshold conservatism and re‑certification effort."
SuTRA (Structurally-Unified Tokenization with Root Awareness) is a morphology-aware tokenization algorithm designed to address the problem of Morphological Shattering in morphologically rich Indic languages. It preserves the indivisibility of aksharas—complex orthographic syllables—by penalizing merges that cross morphological boundaries, thereby reducing over-fragmentation of words. The authors also release a new morphological segmentation dataset for Hindi, Marathi, and Gujarati, and demonstrate that SuTRA improves morphological alignment by up to 14.7% and semantic recoverability by 34% over BPE, leading to an average machine translation gain of +8.08 chrF2.
The paper evaluates open-source OCR, LLM, and VLM systems on a high‑risk public sector task: extracting structured data from student application documents. Results show that VLMs generally outperform OCR+LLM pipelines, yet only 4 of 35 configurations achieve F1 scores above 0.5, with most combinations scoring below 0.25. Model size and input quality, especially preserving OCR structure, are critical factors influencing performance.
The paper presents a local traditional OCR pipeline that can be iteratively fine‑tuned on both layout and appearance levels of complex historical Sanskrit manuscripts. By adapting to the specific manuscript distribution, the pipeline reduces human annotation effort and improves transcription accuracy across subsequent pages. The authors apply this method to three manuscripts, release a dataset with detailed layout and Unicode annotations in PAGE‑XML format, and benchmark the pipeline against leading multimodal large language models.
The paper introduces Structured Prior Knowledge (SPK), a framework that extracts and organizes latent priors from pretrained object detectors to improve out-of-distribution (OoD) detection. SPK uses in-distribution data and hallucination-inducing samples to elicit part-level semantic concepts, then combines these with geometric and contextual priors into a compact five-dimensional representation. Experiments across various detector architectures and OoD benchmarks show that SPK achieves state-of-the-art performance, demonstrating that pretrained detectors encode richer latent knowledge than previously exploited.
The paper introduces a computational framework that detects dynamic team‑process phases in collaborative virtual reality (VR) by analyzing timestamped dialogue. It uses late chunking, penalized Gaussian‑kernel change‑point detection, TF‑IDF, NMF, and a locally deployed large language model to identify semantic transitions and generate interpretable phase descriptions. The detected phases are aligned with interaction logs, demonstrating that transcript‑derived phases correspond to task‑action patterns and thus provide a transparent, transferable method for studying temporal changes in teamwork.
Learning-State-Aware Dynamic Generative Data Augmentation on Small-Scale Datasets proposes LSADA, a method that constructs a learning state for each sample based on its loss and loss‑decrease rate to determine a sample‑specific augmentation strength. LSADA also introduces a decoupled data augmentation and diffusion fusion strategy that applies strength‑controlled transformations to class‑relevant regions while generating diverse class‑irrelevant regions, progressively fusing them to enhance image diversity while preserving class semantics. Experiments on nine public datasets demonstrate that LSADA outperforms the current state‑of‑the‑art dynamic GDA method by an average of 4.5% on six natural image datasets and 2.5% on three medical image datasets.
The paper investigates how the CutMix data augmentation technique affects reliability and robustness in semantic segmentation. It evaluates two architectures—CNN-based DeepLabV3+ and transformer-based SegFormer—on both in-domain and out-of-domain data. Results show that CutMix has a minor effect on segmentation accuracy but consistently improves reliability, especially under distribution shifts, by enhancing calibration and uncertainty quality.