paper-with-me

Papers

Segment-Level Diffusion: A Framework for Controllable Long-Form Generation with Diffusion Language Models

2024-12-15 · Xiaochen Zhu, Georgi Karadzhov, Chenxi Whitehouse, Andreas Vlachos

Diffusion models have shown promise in text generation but often struggle with generating long, coherent, and contextually accurate text. Token-level diffusion overlooks word-order dependencies and enforces short output windows, while passage-level diffusion struggles with learning robust representation for long-form text. To address these challenges, we propose Segment-Level Diffusion (SLD), a framework that enhances diffusion-based text generation through text segmentation, robust representation training with adversarial and contrastive learning, and improved latent-space guidance. By segmenting long-form outputs into separate latent representations and decoding them with an autoregressive decoder, SLD simplifies diffusion predictions and improves scalability. Experiments on XSum, ROCStories, DialogSum, and DeliData demonstrate that SLD achieves competitive or superior performance in fluency, coherence, and contextual compatibility across automatic and human evaluation metrics comparing with other diffusion and autoregressive baselines. Ablation studies further validate the effectiveness of our segmentation and representation learning strategies.

📄 PDF Abstract BibTeX arXiv:2412.11333

Code (0)

등록된 구현이 없습니다.

Tasks

Contrastive LearningDecoderFormRepresentation LearningText GenerationText Segmentation

Methods 이 논문이 사용한 방법론

Diffusion Diffusion models generate samples by gradually removing noise from a signal, and their training objective can be expressed as a reweighted variational lower-bound…

Similar Papers 제목 키워드 기반

SALIENT: Frequency-Aware Paired Diffusion for Controllable Long-Tail CT Detection

2026-02-26 · Yifan Li, Mehrdad Salimitari, Taiyu Zhang, Guang Li 외 arxiv

Detection of rare lesions in whole-body CT is fundamentally limited by extreme class imbalance and low target-to-volume ratios, producing precision collapse despite high AUROC. Synthetic augmentation with diffusion model…

Diffusion Forcing Planner: History-Annealed Planning with Time-Dependent Guidance for Autonomous Driving

2026-06-09 · Zehan Zhang, Neng Zhang, Yaoyi Li, Jia Cai 외 arxiv

Learning-based motion planners, despite recent progress, often suffer from temporal inconsistency. Small perturbations across frames can accumulate into unstable trajectories, degrading comfort and safety in closed-loop …

Autonomous Driving

CLE Diffusion: Controllable Light Enhancement Diffusion Model

2023-08-13 · Yuyang Yin, Dejia Xu, Chuangchuang Tan, Ping Liu 외

Low light enhancement has gained increasing importance with the rapid development of visual creation and editing. However, most existing enhancement algorithms are designed to homogeneously increase the brightness of ima…

Low-Light Image Enhancementmodel

CoSimGen: Controllable Diffusion Model for Simultaneous Image and Mask Generation

2025-03-25 · Rupak Bose, Chinedu Innocent Nwoye, Aditya Bhat, Nicolas Padoy

The acquisition of annotated datasets with paired images and segmentation masks is a critical challenge in domains such as medical imaging, remote sensing, and computer vision. Manual annotation demands significant resou…

Triplet

OmniVDiff: Omni Controllable Video Diffusion for Generation and Understanding

2025-04-15 · Dianbing Xi, Jiepeng Wang, Yuanzhi Liang, Xi Qiu 외

In this paper, we propose a novel framework for controllable video diffusion, OmniVDiff, aiming to synthesize and comprehend multiple video visual content in a single diffusion model. To achieve this, OmniVDiff treats al…

Semantic SegmentationVideo GenerationVideo Understanding