MSN: Multi-Style Network for Trajectory Prediction

Wong, Conghao; Xia, Beihao; Peng, Qinmu; You, Xinge

doi:10.48550/arxiv.2107.00932

Cited by 2 publications

(4 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Tran et al [54] attempt to obtain multi-modal goal proposals from the additional goal channel to generate multiple predictions. Wong et al [33] use a set of generators to give destination proposals with different styles, and then interpolate to forecast multiple future predictions. Others like [53] also introduce several "waypoints" to help better predict agents' potential future intentions rather than the only destination points.…”

Section: Hierarchical Prediction and Bilinear Structuresmentioning

confidence: 99%

“…Spectrums [A, Φ] will query the self-attention feature in the Transformer Decoder to get the final behavior features. We employ another MLP encoder (MLP e ) to aggregate features at different frequency nodes [33], thus inferring agent's behavior feature f . The above process can be written as:…”

Section: -Netmentioning

confidence: 99%

“…Unlike the split of the SDD in TrajNet [88] (which includes only pedestrians), the SDD we use includes all categories of annotations. lines, including Linear Least Square 4 , SoPhie [20], S-BiGAT [19], E-SR-LSTM [85], Multiverse [87], MANTRA [89], TF [28], PECNet [52], STAR [29], Trajectron++ [67], SimAug [86], TPNMS [90], Introvert [91], LB-EBM [92], Agentformer [44], Y-net [53], STC-Net [93], SpecTGNN [77], CSCNet [83], MSN [33], MID [94], SHENet [95], SEEM [96], Social-SSL [66], P-LSTM [40], Zero-Vel [97] , and PV-LSTM [40].…”

Section: Prediction Typesmentioning

confidence: 99%

“…Neural networks like Long-Short Term Memory Networks (LSTMs) [21], [22], [23], Graph Neural Networks (GNNs) [24], [25], [26], [27], and Transformers [28], [29], [30], [31] are employed to encode agents' trajectories. Besides, agents' motion preferences [2], [32], [33] and their goal distributions [34], [35], [36] have been investigated.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

View Vertically: A Hierarchical Network for Trajectory Prediction via Fourier Spectrums

Wong

Xia

Hong

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

With the fast development of AI-related techniques, the applications of trajectory prediction are no longer limited to easier scenes and trajectories. More and more heterogeneous trajectories with different representation forms, such as 2D or 3D coordinates, 2D or 3D bounding boxes, and even high-dimensional human skeletons, need to be analyzed and forecasted. Among these heterogeneous trajectories, interactions between different elements within a frame of trajectory, which we call the "Dimension-Wise Interactions", would be more complex and challenging. However, most previous approaches focus mainly on a specific form of trajectories, which means these methods could not be used to forecast heterogeneous trajectories, not to mention the dimension-wise interaction. Besides, previous methods mostly treat trajectory prediction as a normal time sequence generation task, indicating that these methods may require more work to directly analyze agents' behaviors and social interactions at different temporal scales. In this paper, we bring a new "view" for trajectory prediction to model and forecast trajectories hierarchically according to different frequency portions from the spectral domain to learn to forecast trajectories by considering their frequency responses. Moreover, we try to expand the current trajectory prediction task by introducing the dimension M from "another view", thus extending its application scenarios to heterogeneous trajectories vertically. Finally, we adopt the bilinear structure to fuse two factors, including the frequency response and the dimension-wise interaction, to forecast heterogeneous trajectories via spectrums hierarchically in a generic way. Experiments show that the proposed model outperforms most state-of-the-art methods on ETH-UCY benchmark, Stanford Drone Dataset and nuScenes with heterogeneous trajectories, including 2D coordinates, 2D and 3D bounding boxes. ! * Equal contribution. Codes are available at https://github.com/cocoon2wong/E-Vertical.

show abstract

Section: Hierarchical Prediction and Bilinear Structuresmentioning

confidence: 99%

Section: -Netmentioning

confidence: 99%