Rethinking Training & Inference for Forecasting: Linking Winner-Take-All back to GMMs

• URL: http://arxiv.org/abs/2606.26424v1
• Date: Wed, 24 Jun 2026 22:26:43 GMT
• Title: Rethinking Training & Inference for Forecasting: Linking Winner-Take-All back to GMMs
• Abstract summary: Trajectory forecasting for autonomous driving has advanced rapidly, yet representative models often produce uninformative posteriors over forecast modes.<n>We argue that this K-means-like hard assignment (one-hot) is the source of uninformative mode probabilities.<n>We introduce two post-hoc treatments: (1) test-time posterior-weighted merging that aggregates nearby candidate trajectories; and (2) a one-step expectation-maximization (EM) update.
• Score: 48.245159309501666
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Trajectory forecasting for autonomous driving has advanced rapidly, yet representative models often produce uninformative posteriors over forecast modes, causing problems for mode pruning. We trace this to a modeling-training mismatch: forecasters are typically modeled as conditional Gaussian mixture models (GMMs) but trained with a winner-take-all (WTA) loss that assigns each sample to its nearest mode. We argue that this K-means-like hard assignment (one-hot), while preventing mode collapse, is the source of uninformative mode probabilities: it over-segments the trajectory space, ignores relatedness among nearby modes, and yields assignment instability under small perturbations. Guided by this lens, we introduce two post-hoc treatments: (1) test-time posterior-weighted merging that aggregates nearby candidate trajectories; and (2) a one-step expectation-maximization (EM) update that replaces hard labels with soft responsibilities, sharing probability mass across neighboring modes. Across several WTA-trained architectures, these lightweight steps produce more informative, faithfully ranked mode posteriors and strengthen final forecasts on popular displacement metrics -- without retraining. Our analysis unifies recent design choices through a GMM-vs-K-means perspective and offers principled, practical corrections that better align training objectives with inference.

Related papers

• Unveiling Stochasticity: Universal Multi-modal Probabilistic Modeling for Traffic Forecasting [3.268970214362015]
  This paper proposes an elegant yet universal approach that transforms existing models into probabilistic predictors.<n> Experiments on multiple traffic datasets show that our approach generalizes from classic to modern model architectures.<n>A more detailed study on a real-world dense urban traffic network is presented to examine the impact of data quality uncertainty.
  arXiv  Detail & Related papers  (2026-04-17T14:10:16Z)
• Don't double it: Efficient Agent Prediction in Occlusions [1.1338062042380708]
  Occluded traffic agents pose a significant challenge for autonomous vehicles, as hidden pedestrians or vehicles can appear unexpectedly.<n>Existing learning-based methods often produce redundant occupancy predictions where a single agent is identified multiple times.<n>We introduce MatchInformer, a novel transformer-based approach that builds on the state-of-the-art SceneInformer architecture.
  arXiv  Detail & Related papers  (2026-01-29T10:22:38Z)
• Toward Reliable Machine Unlearning: Theory, Algorithms, and Evaluation [1.7767466724342065]
  We introduce Adrial Machine UNlearning (AMUN), which surpasses prior state-of-the-art methods for image classification based on SOTA MIA scores.<n>We show that existing methods fail in replicating a retrained model's behavior by introducing a nearest-neighbor membership inference attack (MIA-NN)<n>We then propose a fine-tuning objective that mitigates this leakage by approximating, for forget-class inputs, the distribution over remaining classes that a model retrained from scratch would produce.
  arXiv  Detail & Related papers  (2025-12-07T20:57:25Z)
• SynCast: Synergizing Contradictions in Precipitation Nowcasting via Diffusion Sequential Preference Optimization [62.958457694151384]
  We introduce preference optimization into precipitation nowcasting for the first time, motivated by the success of reinforcement learning from human feedback in large language models.<n>In the first stage, the framework focuses on reducing FAR, training the model to effectively suppress false alarms.
  arXiv  Detail & Related papers  (2025-10-22T16:11:22Z)
• Self-Boost via Optimal Retraining: An Analysis via Approximate Message Passing [58.52119063742121]
  Retraining a model using its own predictions together with the original, potentially noisy labels is a well-known strategy for improving the model performance.<n>This paper addresses the question of how to optimally combine the model's predictions and the provided labels.<n>Our main contribution is the derivation of the Bayes optimal aggregator function to combine the current model's predictions and the given labels.
  arXiv  Detail & Related papers  (2025-05-21T07:16:44Z)
• Annealed Winner-Takes-All for Motion Forecasting [48.200282332176094]
  We show how an aWTA loss can be integrated with state-of-the-art motion forecasting models to enhance their performance.<n>Our approach can be easily incorporated into any trajectory prediction model normally trained using WTA.
  arXiv  Detail & Related papers  (2024-09-17T13:26:17Z)
• Towards Motion Forecasting with Real-World Perception Inputs: Are End-to-End Approaches Competitive? [93.10694819127608]
  We propose a unified evaluation pipeline for forecasting methods with real-world perception inputs. Our in-depth study uncovers a substantial performance gap when transitioning from curated to perception-based data.
  arXiv  Detail & Related papers  (2023-06-15T17:03:14Z)
• Distributionally Robust Models with Parametric Likelihood Ratios [123.05074253513935]
  Three simple ideas allow us to train models with DRO using a broader class of parametric likelihood ratios. We find that models trained with the resulting parametric adversaries are consistently more robust to subpopulation shifts when compared to other DRO approaches.
  arXiv  Detail & Related papers  (2022-04-13T12:43:12Z)
• Meta-Learned Confidence for Few-shot Learning [60.6086305523402]
  A popular transductive inference technique for few-shot metric-based approaches, is to update the prototype of each class with the mean of the most confident query examples. We propose to meta-learn the confidence for each query sample, to assign optimal weights to unlabeled queries. We validate our few-shot learning model with meta-learned confidence on four benchmark datasets.
  arXiv  Detail & Related papers  (2020-02-27T10:22:17Z)

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.