Pinhao Song

Peking University

No. 38 Xueyuan Road

Haidian District

Beijing, 100871

China

  • Peking University

SCHOLARLY PAPERS

4

DOWNLOADS

241

TOTAL CITATIONS
Rank 35,037

SSRN RANKINGS

Top 35,037

in Total Papers Citations

4

Scholarly Papers (4)

Actions:
Email selected abstracts
1.

Neural Ordinary Differential Equation for Irregular Human Motion Prediction

Number of pages: 7 Posted: 17 Mar 2023
Yang Chen, Hong Liu, Pinhao Song and Wenhao Li
Peking University - Shenzhen Graduate School, Peking University, Peking University and Peking University
Downloads 121 (507,520)

Abstract:

Loading...

Human motion prediction, Neural ordinary differential equation, Irregular time series

2.

Boosting R-CNN: Reweighting R-CNN Samples by RPN’s Error for Underwater Object Detection

Number of pages: 15 Posted: 04 Oct 2022
Pinhao Song, Pengteng Li, Linhui Dai, Tao Wang and Zhan Chen
Peking University, Shenzhen University, Peking University, Peking University and Peking University
Downloads 68 (736,818)

Abstract:

Loading...

underwater object detection, hard example mining, uncertainty modeling

3.

Achieving Domain Generalization for Underwater Object Detection by Domain Mixup and Contrastive Learning

Number of pages: 14 Posted: 10 Oct 2022
Yang Chen, Hong Liu, Pinhao Song, Linhui Dai, Xiaochuan Zhang, Runwei Ding and Shengquan Li
Peking University - Shenzhen Graduate School, Peking University, Peking University, Peking University, affiliation not provided to SSRN, Peng Cheng Laboratory and Peng Cheng Laboratory
Downloads 34 (1,011,351)

Abstract:

Loading...

Domain Generalizatio, Underwater, Object detection, Image Stylization, contrastive learning

4.

A Gated Cross-Domain Collaborative Network for Underwater Object Detection

Number of pages: 34 Posted: 03 Jul 2023
Linhui Dai, Hong Liu, Pinhao Song and Mengyuan Liu
Peking University, Peking University, Peking University and affiliation not provided to SSRN
Downloads 18 (1,199,945)
Citation 4

Abstract:

Loading...

underwater object detection, underwater image enhancement, cross-domain