TY - GEN
T1 - IDOL
T2 - 18th European Conference on Computer Vision, ECCV 2024
AU - Zhai, Yuanhao
AU - Lin, Kevin
AU - Li, Linjie
AU - Lin, Chung Ching
AU - Wang, Jianfeng
AU - Yang, Zhengyuan
AU - Doermann, David
AU - Yuan, Junsong
AU - Liu, Zicheng
AU - Wang, Lijuan
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
PY - 2025
Y1 - 2025
N2 - Significant advances have been made in human-centric video generation, yet the joint video-depth generation problem remains underexplored. Most existing monocular depth estimation methods may not generalize well to synthesized images or videos, and multi-view-based methods have difficulty controlling the human appearance and motion. In this work, 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, we propose a unified dual-modal U-Net, a parameter-sharing framework for joint video and depth denoising, wherein a modality label guides the denoising target, and cross-modal attention enables the mutual information flow. 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, leading to harmonized outputs. Additionally, a cross-attention map consistency loss is applied to align the cross-attention map of the video denoising with that of the depth denoising, further facilitating spatial alignment. Extensive experiments on the TikTok and NTU120 datasets show our superior performance, significantly surpassing existing methods in terms of video FVD and depth accuracy.
AB - Significant advances have been made in human-centric video generation, yet the joint video-depth generation problem remains underexplored. Most existing monocular depth estimation methods may not generalize well to synthesized images or videos, and multi-view-based methods have difficulty controlling the human appearance and motion. In this work, 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, we propose a unified dual-modal U-Net, a parameter-sharing framework for joint video and depth denoising, wherein a modality label guides the denoising target, and cross-modal attention enables the mutual information flow. 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, leading to harmonized outputs. Additionally, a cross-attention map consistency loss is applied to align the cross-attention map of the video denoising with that of the depth denoising, further facilitating spatial alignment. Extensive experiments on the TikTok and NTU120 datasets show our superior performance, significantly surpassing existing methods in terms of video FVD and depth accuracy.
UR - https://www.scopus.com/pages/publications/85210887751
U2 - 10.1007/978-3-031-72633-0_8
DO - 10.1007/978-3-031-72633-0_8
M3 - Conference contribution
AN - SCOPUS:85210887751
SN - 9783031726323
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 134
EP - 152
BT - Computer Vision – ECCV 2024 - 18th European Conference, Proceedings
A2 - Leonardis, Aleš
A2 - Ricci, Elisa
A2 - Roth, Stefan
A2 - Russakovsky, Olga
A2 - Sattler, Torsten
A2 - Varol, Gül
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 29 September 2024 through 4 October 2024
ER -