Skip to content

MamMA: A Mamba-Based Pedestrian Trajectory Prediction Algorithm Considering Occupancy Map and Pedestrian Awareness States

Sep 2026 · 0 citations · 25 references
Computer Science

TL;DR

MamMA, a Mamba-based pedestrian trajectory prediction algorithm considering occupancy maps and pedestrian awareness states is proposed, which achieves better average displacement error and final displacement error than the state-of-the-art algorithms.

Abstract

Many pedestrian trajectory prediction algorithms have been proposed to improve the safety of navigation for mobile robots working in human-robot coexistence environments. Some pedestrian trajectory prediction algorithms extract information about obstacles near pedestrians from top-down view images to improve the accuracy of trajectory prediction. However, mobile robots typically create local occupancy maps using LiDAR, rather than top-down view images. Meanwhile, the vision sensors on board robots provide egocentric view images, which contain fine-grained behavioral information about the pedestrians near the robot. To better use the information collected by LiDAR and on-board vision sensors, we propose MamMA, a Mamba-based pedestrian trajectory prediction algorithm considering occupancy maps and pedestrian awareness states. MamMA divides the occupancy map by patches and extracts obstacle features from each patch to create map features. Pedestrian awareness states are divided and considered, as some studies show that awareness states affect the perception and speed of pedestrians. Furthermore, a Mamba-based model is proposed to predict the future trajectories of pedestrians based on different types of features. Experiments on the STCrowd, SiT, JRDB, ETH, and UCY datasets show that MamMA achieves better average displacement error and final displacement error than the state-of-the-art algorithms.

View source

Similar papers

Open access Sep 2026

RCN-MAMBA: Research on Pedestrian-Future-Trajectory-Prediction Methods for Occlusion Scenarios

Pedestrian-future-trajectory prediction is a critical task in intelligent driving, traffic-scene understanding, and active safety decision-making. In real-world road environments, pedestrians’ historical trajectories are frequently affected by occlusions from vehicles, other pedestrians, road infrastructure, and detect...

Si-Jie Yang, Zhao-Yu Li, Guo-Yu Lin et al. · 0 citations
Conference Oct 2026

Research on obstacle trajectory prediction strategies based on the dynamic gray model

Obstacle trajectory prediction is a critical task for mobile robots to achieve dynamic obstacle avoidance, with the core challenge lying in enabling robots to accurately and promptly perceive and predict the trajectories of all obstacles within the same scene under varying obstacle motion states. To address this issue,...

Hong-Xia Wang · 0 citations
Open access 2026

Condition-Aware Trajectory Prediction via Velocity-Aware Model Selection

Safe and proactive decision-making in autonomous driving systems depends on predicting the future locations of vehicles, pedestrians, and other dynamic actors in traffic with high precision. However, trajectory prediction is an extremely challenging problem due to the existence of multimodal possible futures, the rando...

Hatice Meltem Nergiz Sirin, O. Erkent · 0 citations
Conference Oct 2026

Vision-based adaptive dynamic path planning for mobile robots in crosswalks

For the problem of real-time navigation and obstacle avoidance of mobile robots in the scenario of a continuously moving pedestrian on a dynamic zebra crossing, this paper proposes a visual-based adaptive dynamic path planning framework. Firstly, the YOLO algorithm is used to detect dynamic pedestrians in the video str...

Jing-Yi Luo · 0 citations
Open access Dec 2026

Human-aware robot navigation in public space based on predicting pedestrian reaction patterns

This paper presents a human-aware collision-free navigation system based on long-term motion prediction and a game against nature formulation for modelling uncertainty. The proposed approach explicitly takes into account mutual pedestrian interactions and introduces a method for estimating the confidence of predicted p...

K. Skrzypczyk · 0 citations
Open access Sep 2026

Systematic Evaluation of Reference Point Selection for Pedestrian Localization in Aerial-Ground Multi-View Monitoring Systems

Accurate pedestrian localization is fundamental to urban monitoring systems such as crowd management and anomalous behavior detection. Multi-view systems combining CCTV and UAV imagery extend monitoring coverage, but pedestrian locations estimated from each domain must be consistent in a common world coordinate system...

Sunwoong Paik, Joonhee Youn, Phillip Kim et al. · 0 citations

Related blog posts

MIT News · Artificial Intelligence Oct 28, 2026

Making machines that can listen

Professor Paris Smaragdis develops computers with superhuman hearing for applications like speech recognition, music understanding, and environmental monitoring.

Microsoft Research Blog Aug 11, 2026

Introducing CARE-X: Towards Clinically Useful Radiology VLMs with Auxiliary Supervision, Reward-Aligned Learning, and Tool-Augmented Measurement

Radiology AI is evolving beyond report generation. CARE-X explores a unified approach that combines flexible reasoning, calibrated predictions, and measurement-based tools for chest X-ray interpretation. The post Introducing CARE-X: Towards Clinically Useful Radiology VLMs with Auxiliary Supervision, Reward-Aligned Learning, and Tool-Augmented Measurement appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.