End-to-end Birds-eye-view Flow Estimation for Autonomous Driving - KH. Lee, M. Kliemann, A. Gaidon, J. Li, C. Fang, S. Pillai, W. Burgard - IROS 2020 - October 25, 2020. An end-to-end deep learning framework for LIDAR-based flow estimation in 2.5D bird’s eye view (BeV). We show it boosts tracking performance on a real-world autonomous car.
Partial list. You can also find my articles on my Google Scholar profile.
It Is Not the Journey but the Destination: Endpoint Conditioned Trajectory Prediction - K. Mangalam, H. Girase, S. Agarwal, K-H. Lee, E. Adeli, J. Malik, A. Gaidon - ECCV 2020 (oral, top 2%) - August 27, 2020. Predicted Endpoint Conditioned Network (PECNet) for flexible human trajectory prediction. PECNet infers distant trajectory endpoints to assist in long-range multi-modal trajectory prediction.
Monocular Differentiable Rendering for Self-Supervised 3D Object Detection - D. Beker, H. Kato, MA. Morariu, T. Ando, T. Matsuoka, W. Kehl, A. Gaidon - ECCV 2020 - August 23, 2020. Novel self-supervised method for textured 3D shape reconstruction and pose estimation of rigid objects with the help of strong shape priors and 2D instance masks.
Neural Ray Surfaces for Self-Supervised Learning of Depth and Ego-motion - I. Vasiljevic, V. Guizilini, R. Ambrus, S. Pillai, W. Burgard, G. Shakhnarovich, A. Gaidon - arxiv 2020 - August 15, 2020. Neural Ray Surfaces (NRS) are convolutional networks that represent pixel-wise projection rays, approximating a wide range of cameras. NRS are fully differentiable and can be learned end-to-end from unlabeled raw videos.
Reinforcement Learning based Control of Imitative Policies for Near-Accident Driving - Z. Cao, E. Biyik, W. Z. Wang, A. Raventos, A. Gaidon, G. Rosman, D. Sadigh - RSS 2020 - July 15, 2020. We propose a hierarchical reinforcement and imitation learning (H-ReIL) approach for learning to drive in near-accident scenarios.
Differentiable Rendering: A Survey - H. Kato, D. Beker, M. Morariu, T. Ando, T. Matsuoka, W. Kehl, A. Gaidon - arXiv - June 22, 2020. This paper reviews existing literature and discusses the current state of differentiable rendering, its applications, and open research problems.
Spatio-Temporal Graph for Video Captioning with Knowledge Distillation - B. Pan, H. Cai, DA Huang, KH Lee, A. Gaidon, E. Adeli, JC Niebles - CVPR 2020 - June 16, 2020. We propose a model that learns to distill spatio-temporal object interactions for video captioning.
Real-Time Panoptic Segmentation from Dense Detections - R. Hou, J. Li, A. Bhargava, A. Raventos, V. Guizilini, C. Fang, J Lynch, A. Gaidon - CVPR 2020 (oral, top 5.7%) - June 16, 2020. A novel panoptic segmentation method featuring parameter-free instance mask reconstruction, state-of-the-art accuracy, and real-time inference.
Autolabeling 3D Objects with Differentiable Rendering of SDF Shape Priors” - S. Zakharov, W. Kehl, A. Bhargava, A. Gaidon - CVPR 2020 (oral, top 5.7%) - June 16, 2020. We present an automatic annotation pipeline to recover 9D cuboids and 3D shapes from pre-trained off-the-shelf 2D detectors and sparse LIDAR data.
3D Packing for Self-Supervised Monocular Depth Estimation - V. Guizilini, R. Ambrus, S. Pillai, A. Raventos, A. Gaidon - CVPR 2020 (oral, top 5.7%) - June 16, 2020. We introduce a new state of the art in self-supervised monocular depth estimation (PackNet), a new benchmark (DDAD), and weak velocity supervision.
Spatiotemporal Relationship Reasoning for Pedestrian Intent Prediction - B. Liu, E. Adeli, Z. Cao, KH Lee, A. Shenoi, A. Gaidon, JC Niebles - RA-L & ICRA 2020 - May 31, 2020. We propose a graph convolutional model operating on scene graphs to predict pedestrian crossing intent and introduce the STIP dataset.
Semantically-Guided Representation Learning for Self-Supervised Monocular Depth - V. Guizilini, R. Hou, J. Li, R. Ambrus, A. Gaidon - ICLR 2020 - April 26, 2020. A new architecture leveraging fixed pretrained semantic segmentation networks to guide self-supervised representation learning via pixel-adaptive convolutions.