Task-oriented Learnable Diffusion Timesteps for Universal Few-shot Learning of Dense Tasks

• URL: http://arxiv.org/abs/2512.23210v2
• Date: Wed, 31 Dec 2025 21:09:53 GMT
• Title: Task-oriented Learnable Diffusion Timesteps for Universal Few-shot Learning of Dense Tasks
• Authors: Changgyoon Oh, Jongoh Jeong, Jegyeong Cho, Kuk-Jin Yoon,
• Abstract summary: Current diffusion model-based applications exploit the power of learned visual representations from multistep forward-backward Markovian processes for single-task prediction tasks.<n>We propose two modules: Task-aware Timestep Selection (TTS) to select ideal diffusion timesteps based on timestep-wise losses and similarity scores, and Timestep Feature Consolidation (TFC) to consolidate the selected timestep features.<n>Our framework effectively achieves superiority in dense prediction performance given only a few support queries.
• Score: 48.86985692711283
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Denoising diffusion probabilistic models have brought tremendous advances in generative tasks, achieving state-of-the-art performance thus far. Current diffusion model-based applications exploit the power of learned visual representations from multistep forward-backward Markovian processes for single-task prediction tasks by attaching a task-specific decoder. However, the heuristic selection of diffusion timestep features still heavily relies on empirical intuition, often leading to sub-optimal performance biased towards certain tasks. To alleviate this constraint, we investigate the significance of versatile diffusion timestep features by adaptively selecting timesteps best suited for the few-shot dense prediction task, evaluated on an arbitrary unseen task. To this end, we propose two modules: Task-aware Timestep Selection (TTS) to select ideal diffusion timesteps based on timestep-wise losses and similarity scores, and Timestep Feature Consolidation (TFC) to consolidate the selected timestep features to improve the dense predictive performance in a few-shot setting. Accompanied by our parameter-efficient fine-tuning adapter, our framework effectively achieves superiority in dense prediction performance given only a few support queries. We empirically validate our learnable timestep consolidation method on the large-scale challenging Taskonomy dataset for dense prediction, particularly for practical universal and few-shot learning scenarios.

Related papers

• Streaming Real-Time Trajectory Prediction Using Endpoint-Aware Modeling [54.94692733670454]
  Future trajectories of neighboring traffic agents have a significant influence on the path planning and decision-making of autonomous vehicles.<n>We propose a lightweight yet highly accurate streaming-based trajectory forecasting approach.<n>Our approach significantly reduces inference latency, making it well-suited for real-world deployment.
  arXiv  Detail & Related papers  (2026-03-02T13:44:23Z)
• Multi-Scale Finetuning for Encoder-based Time Series Foundation Models [67.95907033226585]
  Time series foundation models (TSFMs) demonstrate impressive zero-shot performance for time series forecasting.<n>While naive finetuning can yield performance gains, we argue that it falls short of fully leveraging TSFMs' capabilities.<n>We propose Multiscale finetuning (MSFT), a simple yet general framework that explicitly integrates multi-scale modeling into the finetuning process.
  arXiv  Detail & Related papers  (2025-06-17T01:06:01Z)
• Winner-takes-all for Multivariate Probabilistic Time Series Forecasting [0.0]
  We introduce TimeMCL, a method leveraging the Multiple Choice Learning (MCL) paradigm to forecast plausible time series futures.<n>Our approach employs a neural network with multiple heads and utilizes the Winner-Takes-All (WTA) loss to promote diversity among predictions.
  arXiv  Detail & Related papers  (2025-06-05T18:56:14Z)
• Task-Agnostic Pre-training and Task-Guided Fine-tuning for Versatile Diffusion Planner [12.360598915420255]
  Diffusion models have demonstrated their capabilities in modeling trajectories of multi-tasks.<n>Existing multi-task planners or policies typically rely on task-specific demonstrations via multi-task imitation, or require task-specific reward labels.<n>We propose a versatile diffusion planner capable of leveraging large-scale inferior data that contains task-agnostic sub-optimal trajectories.
  arXiv  Detail & Related papers  (2024-09-30T05:05:37Z)
• Align Your Steps: Optimizing Sampling Schedules in Diffusion Models [63.927438959502226]
  Diffusion models (DMs) have established themselves as the state-of-the-art generative modeling approach in the visual domain and beyond. A crucial drawback of DMs is their slow sampling speed, relying on many sequential function evaluations through large neural networks. We propose a general and principled approach to optimizing the sampling schedules of DMs for high-quality outputs.
  arXiv  Detail & Related papers  (2024-04-22T18:18:41Z)
• Spurious Feature Eraser: Stabilizing Test-Time Adaptation for Vision-Language Foundation Model [86.9619638550683]
  Vision-language foundation models have exhibited remarkable success across a multitude of downstream tasks due to their scalability on extensive image-text paired data.<n>However, these models display significant limitations when applied to downstream tasks, such as fine-grained image classification, as a result of decision shortcuts''
  arXiv  Detail & Related papers  (2024-03-01T09:01:53Z)
• Distillation Enhanced Time Series Forecasting Network with Momentum Contrastive Learning [7.4106801792345705]
  We propose DE-TSMCL, an innovative distillation enhanced framework for long sequence time series forecasting. Specifically, we design a learnable data augmentation mechanism which adaptively learns whether to mask a timestamp. Then, we propose a contrastive learning task with momentum update to explore inter-sample and intra-temporal correlations of time series. By developing model loss from multiple tasks, we can learn effective representations for downstream forecasting task.
  arXiv  Detail & Related papers  (2024-01-31T12:52:10Z)
• Successive Model-Agnostic Meta-Learning for Few-Shot Fault Time Series Prognosis [3.5573601621032944]
  We introduce a novel 'pseudo meta-task' partitioning scheme that treats a continuous time period of a time series as a meta-task. Employing continuous time series as pseudo meta-tasks allows our method to extract more comprehensive features and relationships from the data. We introduce a differential algorithm to enhance the robustness of our method across different datasets.
  arXiv  Detail & Related papers  (2023-11-04T02:07:47Z)
• Generative Time Series Forecasting with Diffusion, Denoise, and Disentanglement [51.55157852647306]
  Time series forecasting has been a widely explored task of great importance in many applications. It is common that real-world time series data are recorded in a short time period, which results in a big gap between the deep model and the limited and noisy time series. We propose to address the time series forecasting problem with generative modeling and propose a bidirectional variational auto-encoder equipped with diffusion, denoise, and disentanglement.
  arXiv  Detail & Related papers  (2023-01-08T12:20:46Z)
• Temporally Correlated Task Scheduling for Sequence Learning [143.70523777803723]
  In many applications, a sequence learning task is usually associated with multiple temporally correlated auxiliary tasks. We introduce a learnable scheduler to sequence learning, which can adaptively select auxiliary tasks for training. Our method significantly improves the performance of simultaneous machine translation and stock trend forecasting.
  arXiv  Detail & Related papers  (2020-07-10T10:28:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.