3940 papers
Pinterest introduces Canvas, a large-scale image generation system that employs a foundational diffusion model rapidly fine-tuned into specialized variants for specific editing tasks, achieving significant engagement lifts and outperforming third-party models in production A/B tests.
This paper systematically investigates how loss weighting and output parameterization choices in flow matching models interact with data dimensionality, architecture, and dataset size to provide practical design insights for optimizing both denoising accuracy and generative quality.
This paper demonstrates that frozen vision-language model features contain rich, continuous geometric information that outperforms text-based outputs by 3.3x, revealing that the accuracy bottleneck stems from training objectives and autoregressive generation rather than representational limitations, as evidenced by high-precision linear probes and consistent performance across diverse encoder architectures.
GreenRFM introduces a resource-efficient pre-training framework utilizing principled "MUST" supervision to achieve state-of-the-art radiology foundation model performance with significantly reduced computational requirements, challenging the prevailing "scale is all you need" paradigm.
Match4Annotate is a lightweight framework that enables efficient, high-quality propagation of sparse point and mask annotations across and within video sequences by fitting test-time implicit neural representations to DINOv3 features, offering a scalable solution for annotation bottlenecks in specialized domains like medical imaging.
This paper introduces Self-Flow, a self-supervised flow matching paradigm that utilizes a Dual-Timestep Scheduling mechanism to integrate representation learning directly into the generative framework, thereby eliminating the need for external models and achieving superior, scalable multi-modal synthesis across image, video, and audio.
This paper introduces SG-DOR, a relational framework that utilizes a direction-aware graph neural network to infer scene graphs encoding physical attachments and direction-conditioned occlusion for robotic harvesting of pepper plants in dense canopies.
This paper presents an artificial intelligence system trained on over 45,000 ultrasound images that achieves diagnostic accuracy comparable to senior radiologists for fetal orofacial clefts, significantly enhances junior radiologists' performance when used as a copilot, and accelerates clinical expertise development for rare conditions.
This paper introduces SCAN, a universal framework that leverages an AutoEncoder-based approach guided by the Information Bottleneck principle to generate high-resolution, faithful self-confidence maps, effectively overcoming the trade-off between fidelity and applicability in visual explanations for both CNN and Transformer architectures.
The paper proposes AV-Unified, a novel framework that unifies diverse audio-visual scene understanding tasks into a single architecture by converting heterogeneous inputs and outputs into discrete token sequences and employing multi-scale spatiotemporal perception modules to effectively capture cross-modal associations.
This paper introduces a spatial calibration method for diffuse direct time-of-flight LiDARs that estimates per-pixel footprints and sensitivity maps in an RGB image plane using a scanned retroreflective patch, thereby enabling accurate cross-modal alignment and fusion despite the violation of standard single-ray assumptions.
This paper introduces NEGATE, a training-free framework that models linguistic negation in text-to-video diffusion as a structured feasibility constraint, enabling robust and coherent generation of negated concepts by projecting semantic updates onto a convex set derived from linguistic structure without retraining the underlying models.
The paper introduces SurgFormer, a scalable multiresolution gated transformer that enables near real-time, high-fidelity soft tissue simulation on volumetric meshes by learning to predict nodewise displacements and handling topology-altering resections through a unified, XFEM-supervised framework.
This paper investigates redundancy as a critical yet underexplored data quality factor in autonomous driving by modeling and measuring it across multisource and multimodal datasets, demonstrating that selectively removing redundant labels from overlapping camera views and image-LiDAR pairs can improve or maintain object detection performance while advocating for a data-centric approach to AV dataset optimization.
The paper introduces EgoReasoner, a two-stage framework that employs task-adaptive thinking templates and task-aware reinforcement learning to overcome the limitations of generic reasoning methods, enabling a compact 3B-parameter model to significantly outperform larger vision-language models on complex egocentric 4D reasoning tasks.
Penguin-VL challenges the reliance on massive contrastive pretraining for vision encoders by introducing an LLM-initialized encoder that achieves superior performance in fine-grained perception and complex reasoning tasks with compact, compute-efficient models.
The paper introduces SUREON, a large-scale surgical video QA dataset derived from academic lecture narrations, and presents two specialized vision-language models that demonstrate superior surgical reasoning capabilities by explicitly interpreting intent, rationale, and future steps in surgical scenes.
SCOPE introduces a plug-and-play framework for incremental few-shot 3D segmentation that enriches novel class prototypes by retrieving and fusing high-confidence pseudo-instances from unlabelled background regions, thereby achieving state-of-the-art performance on ScanNet and S3DIS while mitigating catastrophic forgetting without retraining the backbone.
The paper proposes BEVLM, a framework that bridges the gap between spatially consistent Bird's-Eye View representations and Large Language Models by distilling semantic knowledge, thereby significantly enhancing both cross-view reasoning accuracy and safety-critical end-to-end driving performance.
Omni-Diffusion introduces the first any-to-any multimodal language model that unifies text, speech, and image understanding and generation by leveraging a novel mask-based discrete diffusion architecture, demonstrating performance comparable to or exceeding existing autoregressive multimodal systems.