Recent 3D content generation pipelines commonly employ Variational Autoencoders (VAEs) to encode shapes into compact latent representations for diffusion-based generation. However, the widely adopted uniform point sampling strategy in Shape VAE training often leads to a significant loss of geometric details, limiting the quality of shape reconstruction and downstream generation tasks. We present Dora-Vae, a novel approach that enhances VAE reconstruction through our proposed sharp edge sampling strategy and a dual cross-attention mechanism. By identifying and prioritizing regions with high geometric complexity during training, our method significantly improves the preservation of fine-grained shape features. Such sampling strategy and the dual attention mechanism enable the VAE to focus on crucial geometric details that are typically missed by uniform sampling approaches. To systematically evaluate VAE reconstruction quality, we additionally propose Dora-Bench, a benchmark that quantifies shape complexity through the density of sharp edges, introducing a new metric focused on reconstruction accuracy at these salient geometric features. Extensive experiments on the Dora-Bench demonstrate that Dora-Vae achieves comparable reconstruction quality to the state-of-the-art dense XCube-Vae while requiring a latent space at least 8× smaller (1,280 vs. > 10,000 codes). Project page: https://aruichen.github.io/Dora.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Dora: Sampling and Benchmarking for 3D Shape Variational Auto-Encoders


    Beteiligte:
    Chen, Rui (Autor:in) / Zhang, Jianfeng (Autor:in) / Liang, Yixun (Autor:in) / Luo, Guan (Autor:in) / Li, Weiyu (Autor:in) / Liu, Jiarui (Autor:in) / Li, Xiu (Autor:in) / Long, Xiaoxiao (Autor:in) / Feng, Jiashi (Autor:in) / Tan, Ping (Autor:in)


    Erscheinungsdatum :

    10.06.2025


    Format / Umfang :

    9074506 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Dora: Sampling and Benchmarking for 3D Shape Variational Auto-Encoders

    Chen, Rui / Zhang, Jianfeng / Liang, Yixun et al. | ArXiv | 2024

    Freier Zugriff

    Consistency Regularization for Variational Auto-Encoders

    Sinha, Samarth / Dieng, Adji B. | ArXiv | 2021

    Freier Zugriff

    Variational auto-encoders with Student's t-prior

    Abiri, Najmeh / Ohlsson, Mattias | ArXiv | 2020

    Freier Zugriff

    Conditional Image Generation by Conditioning Variational Auto-Encoders

    Harvey, William / Naderiparizi, Saeid / Wood, Frank | ArXiv | 2021

    Freier Zugriff