A training-free method for multi-modal trajectory prediction that achieves comparable accuracy to a 57M-parameter transformer while requiring no GPU and zero learned parameters, which enables deployment in new geographic regions from an order of magnitude less historical data.
Abstract
We present a training-free method for multi-modal trajectory prediction that achieves comparable accuracy to a 57M-parameter transformer while requiring no GPU and zero learned parameters. The method builds a transition table of historical state-to-next position pairs and retrieves neighbors using a product kernel over spatial proximity, bearing, speed, and temporal context. Two inference modes operate over this shared representation: diversity-penalized sampling produces trajectories covering distinct plausible routes, while beam search finds the highest-likelihood path. On the TrAISformer benchmark (Danish Maritime AIS), our method achieves competitive accuracy at full data availability and dramatically outperforms the transformer in data-scarce regimes---remaining stable down to 10% of training data where TrAISformer degrades catastrophically. This enables deployment in new geographic regions from an order of magnitude less historical data.
This thesis demonstrates that moderate-degree polynomials capture real-world motion dynamics with high fidelity without constraining predictive performance, and shows that standard in-distribution evaluation and regression-based metrics may fail to reflect true model generalization and prediction plausibility.
This work proposes a new algorithm called Acceleration Matching, which consists of lifting the original interpolation problem to phase space and then regressing onto an explicit conditional acceleration field that induces random, smooth trajectories that agree with the prescribed marginals.
Bartolo Dazzini, Giovanni Conforti, Alain Durmus et al.· 1 citation
Trajectory prediction is essential for many robotic applications, yet most existing models rely on fixed-length observations and struggle with temporally irregular inputs. In real-world settings, prediction difficulty further increases when agents exhibit strong maneuverability, as their future motions depend on distin...
Shuobo Wang, Wen-Yuan Qin, Yong-Zhao Hua et al.· IEEE Robotics and Automation...· 0 citations
The Conditional Informer is proposed, a novel encoder-decoder architecture that formulates trajectory prediction as a conditional generation task that outperforms kinematic and concatenation-based baselines by 15.4% in prediction accuracy when context is available.
Yuansheng Guan, C. Squires, Timothy Hu et al.· arXiv.org· 0 citations
Visual localization is vital for autonomous systems but remains challenging under dynamic conditions. Transformers offer strong temporal modeling at quadratic cost, while CNNs are efficient yet limited in long-range dependencies. Existing methods also lack robustness to illumination, weather, and seasonal changes, cons...
Zhenyu Li, Tian-Yi Shang· IEEE Transactions on Automat...· 0 citations
The increasing availability of trajectory data is often hampered by sparsity and noise. Existing trajectory recovery methods are further limited by either information loss from coordinate discretization into location IDs, or the inefficiency of conventional diffusion models that require a lengthy denoising process from...
Jia-Qi Duan, Shengwei Tian, Long Yu et al.· Proceedings of the Thirty-Fi...· 0 citations
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduOct 14, 2026
Learn how Google DeepMind and GovTech Singapore uses lightweight on-device Gemma models to intercept conversational scams in private chats with zero cloud data egress.
MIT News · Artificial Intelligence· news.mit.eduOct 13, 2026
Machine-learning networks can learn without detailed instructions from a designer. Those that do may offer insight into how the brain reorganizes itself during learning.
