arXiv:2605.29583v1 Announce Type: new 
Abstract: High-capacity watermarking is necessary for 3D Gaussian Splatting (3DGS) assets to embed rich information (e.g., ownership, provenance, and authentication codes), enabling reliable identification and integrity verification in large-scale 3D asset pipelines. Existing bit-to-token watermarking methods based on a pre-trained text encoder are limited to 77-bit messages due to CLIP's fixed 77-token context length, as tokens beyond this limit are unsupported by learned positional embeddings. To address this limitation, we introduce BitC-3DGS, a bit-compression framework that encodes multiple message bits per token. It employs a bit-compressed tokenization scheme that encodes multiple bits within the same chunk into a single semantic token. To enable recovery of the compressed information, it further introduces a dual-branch architecture for joint chunk decompression and bit decoding, along with a hard-message sampling strategy to improve combinatorial coverage during decoder training. Extensive experiments on the Blender and LLFF datasets demonstrate the effectiveness of BitC-3DGS for high-capacity watermarking, achieving high message recovery accuracy and rendering fidelity. For example, it supports 128-bit message capacity with recovery accuracy comparable to that of 64-bit messages in recent state-of-the-art methods.

تم تقديم إطار جديد يسمى BitC-3DGS لتحسين العلامات المائية عالية السعة لأصول 3D Gaussian Splatting (3DGS)، مما يسمح بإدماج معلومات غنية مثل الملكية وأكواد المصادقة. تعالج هذه التطورات القيود المفروضة على الأساليب الحالية التي تقتصر على رسائل تبلغ 77 بت بسبب أطوال سياق الرموز الثابتة في مشفرات النص المدربة مسبقًا مثل CLIP.

Se ha introducido un nuevo marco llamado BitC-3DGS para mejorar el marcado de alta capacidad en activos de 3D Gaussian Splatting (3DGS), permitiendo la incorporación de información extensa como la propiedad y códigos de autenticación. Este avance aborda las limitaciones de los métodos existentes que están restringidos a mensajes de 77 bits debido a las longitudes de contexto de tokens fijas en codificadores de texto preentrenados como CLIP.

Un nouveau cadre nommé BitC-3DGS a été introduit pour améliorer le marquage à haute capacité des actifs de 3D Gaussian Splatting (3DGS), permettant l'intégration d'informations étendues telles que la propriété et les codes d'authentification. Cette avancée répond aux limitations des méthodes existantes qui sont restreintes à des messages de 77 bits en raison des longueurs de contexte de jetons fixes dans les encodeurs de texte pré-entraînés comme CLIP.

A new framework named BitC-3DGS has been introduced to enhance high-capacity watermarking for 3D Gaussian Splatting (3DGS) assets, allowing for the embedding of extensive information such as ownership and authentication codes. This advancement addresses the limitations of existing methods that are restricted to 77-bit messages due to fixed token context lengths in pre-trained text encoders like CLIP.