DualDiff+: Dual-Branch Diffusion for High-Fidelity Video Generation with Reward Guidance

• URL: http://arxiv.org/abs/2503.03689v1
• Date: Wed, 05 Mar 2025 17:31:45 GMT
• Title: DualDiff+: Dual-Branch Diffusion for High-Fidelity Video Generation with Reward Guidance
• Authors: Zhao Yang, Zezhong Qian, Xiaofan Li, Weixiang Xu, Gongpeng Zhao, Ruohong Yu, Lingsi Zhu, Longjun Liu,
• Abstract summary: We present DualDiff, a conditional diffusion model designed to enhance driving scene generation across multiple views and video sequences.<n>To improve synthesis of fine-grained foreground objects, we propose a Foreground-Aware Mask (FGM) denoising loss function.<n>We also develop the Semantic Fusion Attention (SFA) mechanism to dynamically prioritize relevant information and suppress noise.
• Score: 5.113012982922924
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Accurate and high-fidelity driving scene reconstruction demands the effective utilization of comprehensive scene information as conditional inputs. Existing methods predominantly rely on 3D bounding boxes and BEV road maps for foreground and background control, which fail to capture the full complexity of driving scenes and adequately integrate multimodal information. In this work, we present DualDiff, a dual-branch conditional diffusion model designed to enhance driving scene generation across multiple views and video sequences. Specifically, we introduce Occupancy Ray-shape Sampling (ORS) as a conditional input, offering rich foreground and background semantics alongside 3D spatial geometry to precisely control the generation of both elements. To improve the synthesis of fine-grained foreground objects, particularly complex and distant ones, we propose a Foreground-Aware Mask (FGM) denoising loss function. Additionally, we develop the Semantic Fusion Attention (SFA) mechanism to dynamically prioritize relevant information and suppress noise, enabling more effective multimodal fusion. Finally, to ensure high-quality image-to-video generation, we introduce the Reward-Guided Diffusion (RGD) framework, which maintains global consistency and semantic coherence in generated videos. Extensive experiments demonstrate that DualDiff achieves state-of-the-art (SOTA) performance across multiple datasets. On the NuScenes dataset, DualDiff reduces the FID score by 4.09% compared to the best baseline. In downstream tasks, such as BEV segmentation, our method improves vehicle mIoU by 4.50% and road mIoU by 1.70%, while in BEV 3D object detection, the foreground mAP increases by 1.46%. Code will be made available at https://github.com/yangzhaojason/DualDiff.

Related papers

• LargeAD: Large-Scale Cross-Sensor Data Pretraining for Autonomous Driving [52.83707400688378]
  LargeAD is a versatile and scalable framework designed for large-scale 3D pretraining across diverse real-world driving datasets.<n>Our framework leverages VFMs to extract semantically rich superpixels from 2D images, which are aligned with LiDAR point clouds to generate high-quality contrastive samples.<n>Our approach delivers significant performance improvements over state-of-the-art methods in both linear probing and fine-tuning tasks for both LiDAR-based segmentation and object detection.
  arXiv  Detail & Related papers  (2025-01-07T18:59:59Z)
• Progressive Multi-Modal Fusion for Robust 3D Object Detection [12.048303829428452]
  Existing methods perform sensor fusion in a single view by projecting features from both modalities either in Bird's Eye View (BEV) or Perspective View (PV) We propose ProFusion3D, a progressive fusion framework that combines features in both BEV and PV at both intermediate and object query levels. Our architecture hierarchically fuses local and global features, enhancing the robustness of 3D object detection.
  arXiv  Detail & Related papers  (2024-10-09T22:57:47Z)
• IDOL: Unified Dual-Modal Latent Diffusion for Human-Centric Joint Video-Depth Generation [136.5813547244979]
  We present IDOL (unIfied Dual-mOdal Latent diffusion) for high-quality human-centric joint video-depth generation. Our IDOL consists of two novel designs. First, to enable dual-modal generation and maximize the information exchange between video and depth generation. Second, to ensure a precise video-depth spatial alignment, we propose a motion consistency loss that enforces consistency between the video and depth feature motion fields.
  arXiv  Detail & Related papers  (2024-07-15T17:36:54Z)
• E2E-MFD: Towards End-to-End Synchronous Multimodal Fusion Detection [21.185032466325737]
  We introduce E2E-MFD, a novel end-to-end algorithm for multimodal fusion detection.<n>E2E-MFD streamlines the process, achieving high performance with a single training phase.<n>Our extensive testing on multiple public datasets reveals E2E-MFD's superior capabilities.
  arXiv  Detail & Related papers  (2024-03-14T12:12:17Z)
• UniM$^2$AE: Multi-modal Masked Autoencoders with Unified 3D Representation for 3D Perception in Autonomous Driving [47.590099762244535]
  Masked Autoencoders (MAE) play a pivotal role in learning potent representations, delivering outstanding results across various 3D perception tasks. This research delves into multi-modal Masked Autoencoders tailored for a unified representation space in autonomous driving. To intricately marry the semantics inherent in images with the geometric intricacies of LiDAR point clouds, we propose UniM$2$AE.
  arXiv  Detail & Related papers  (2023-08-21T02:13:40Z)
• M$^2$-3DLaneNet: Exploring Multi-Modal 3D Lane Detection [30.250833348463633]
  M$2$-3DLaneNet lifts 2D features into 3D space by incorporating geometry information from LiDAR data through depth completion. Experiments on the large-scale OpenLane dataset demonstrate the effectiveness of M$2$-3DLaneNet, regardless of the range.
  arXiv  Detail & Related papers  (2022-09-13T13:45:18Z)
• AutoAlignV2: Deformable Feature Aggregation for Dynamic Multi-Modal 3D Object Detection [17.526914782562528]
  We propose AutoAlignV2, a faster and stronger multi-modal 3D detection framework, built on top of AutoAlign. Our best model reaches 72.4 NDS on nuScenes test leaderboard, achieving new state-of-the-art results.
  arXiv  Detail & Related papers  (2022-07-21T06:17:23Z)
• BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird's-Eye View Representation [105.96557764248846]
  We introduce BEVFusion, a generic multi-task multi-sensor fusion framework. It unifies multi-modal features in the shared bird's-eye view representation space. It achieves 1.3% higher mAP and NDS on 3D object detection and 13.6% higher mIoU on BEV map segmentation, with 1.9x lower cost.
  arXiv  Detail & Related papers  (2022-05-26T17:59:35Z)
• DSGN++: Exploiting Visual-Spatial Relation forStereo-based 3D Detectors [60.88824519770208]
  Camera-based 3D object detectors are welcome due to their wider deployment and lower price than LiDAR sensors. We revisit the prior stereo modeling DSGN about the stereo volume constructions for representing both 3D geometry and semantics. We propose our approach, DSGN++, aiming for improving information flow throughout the 2D-to-3D pipeline.
  arXiv  Detail & Related papers  (2022-04-06T18:43:54Z)
• DeepFusion: Lidar-Camera Deep Fusion for Multi-Modal 3D Object Detection [83.18142309597984]
  Lidars and cameras are critical sensors that provide complementary information for 3D detection in autonomous driving. We develop a family of generic multi-modal 3D detection models named DeepFusion, which is more accurate than previous methods.
  arXiv  Detail & Related papers  (2022-03-15T18:46:06Z)
• Dense Voxel Fusion for 3D Object Detection [10.717415797194896]
  Voxel Fusion (DVF) is a sequential fusion method that generates multi-scale dense voxel feature representations. We train directly with ground truth 2D bounding box labels, avoiding noisy, detector-specific, 2D predictions. We show that our proposed multi-modal training strategy results in better generalization compared to training using erroneous 2D predictions.
  arXiv  Detail & Related papers  (2022-03-02T04:51:31Z)
• EPMF: Efficient Perception-aware Multi-sensor Fusion for 3D Semantic Segmentation [62.210091681352914]
  We study multi-sensor fusion for 3D semantic segmentation for many applications, such as autonomous driving and robotics. In this work, we investigate a collaborative fusion scheme called perception-aware multi-sensor fusion (PMF) We propose a two-stream network to extract features from the two modalities separately. The extracted features are fused by effective residual-based fusion modules.
  arXiv  Detail & Related papers  (2021-06-21T10:47:26Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.