Abstract:Temporal motion modeling has always been a key component in multiple object tracking (MOT) which can ensure smooth trajectory movement and provide accurate positional information to enhance association precision. However, current motion models struggle to be both efficient and effective across different application scenarios. To this end, we propose TrackSSM inspired by the recently popular state space models (SSM), a unified encoder-decoder motion framework that uses data-dependent state space model to perform temporal motion of trajectories. Specifically, we propose Flow-SSM, a module that utilizes the position and motion information from historical trajectories to guide the temporal state transition of object bounding boxes. Based on Flow-SSM, we design a flow decoder. It is composed of a cascaded motion decoding module employing Flow-SSM, which can use the encoded flow information to complete the temporal position prediction of trajectories. Additionally, we propose a Step-by-Step Linear (S$^2$L) training strategy. By performing linear interpolation between the positions of the object in the previous frame and the current frame, we construct the pseudo labels of step-by-step linear training, ensuring that the trajectory flow information can better guide the object bounding box in completing temporal transitions. TrackSSM utilizes a simple Mamba-Block to build a motion encoder for historical trajectories, forming a temporal motion model with an encoder-decoder structure in conjunction with the flow decoder. TrackSSM is applicable to various tracking scenarios and achieves excellent tracking performance across multiple benchmarks, further extending the potential of SSM-like temporal motion models in multi-object tracking tasks. Code and models are publicly available at \url{<a class="link-external link-https" href="https://github.com/Xavier-Lin/TrackSSM" rel="external noopener nofollow">this https URL</a>}.

Shared state space model for background information extraction and time series prediction

Relational State-Space Model for Stochastic Multi-Object Systems

Effectively Modeling Time Series with Simple Discrete State Spaces

Time-SSM: Simplifying and Unifying State Space Models for Time Series Forecasting

Kalman-SSM: Modeling Long-Term Time Series With Kalman Filter Structured State Spaces

Research on Deep Spatio-Temporal Model and Its Application in Situation Prediction

State Space Models on Temporal Graphs: A First-Principles Study

Spectral State Space Models

TrackSSM: A General Motion Predictor by State-Space Model

On Understanding of Spatiotemporal Prediction Model

State-Sharing Sparse Hidden Markov Models for Personalized Sequences

Hierarchical State Space Models for Continuous Sequence-to-Sequence Modeling

Predicting High-Dimensional Time Series Data with Spatial, Temporal and Global Information

Coupling LSTM neural networks and state-space models through analytically tractable inference

State-space Decomposition Model for Video Prediction Considering Long-term Motion Trend

SSDNet: State Space Decomposition Neural Network for Time Series Forecasting

Learning Interpretable Deep State Space Model for Probabilistic Time Series Forecasting

State Space Gaussian Process Prediction

SpikingSSMs: Learning Long Sequences with Sparse and Parallel Spiking State Space Models

Is Single Enough? A Joint Spatiotemporal Feature Learning Framework for Multivariate Time Series Prediction

Time Series Analysis by State Space Learning