Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.10343

Real-Time Joint Audio-Video Generation by Parallel Adapter Composition

Jingyu Li·Xiaoxiao Xiang·Yiwen Guo
View original source ↗Read source document ↗Export metadata ↓

Abstract

Deploying a joint audio-video diffusion transformer for real-time, interactive generation normally requires two essential modifications: block-autoregressive attention, so frames can be emitted before the whole clip is finished, and few-step sampling, so each block is cheap. Conventionally, the streaming video literature obtains both capabilities from a chained pipeline. It first distills a bidirectional teacher into a causal student, then into a few-step one, or proceeds in reverse order. Each stage of such a chain fine-tunes the weights the previous one produced, so a later objective can undo an earlier capability. Following the idea of model merging, we show that on a packed audio-video backbone the two capabilities can be acquired in parallel. A causal adapter is trained against the frozen backbone, and an off-the-shelf few-step adapter provides the few-step capability. As the two edit different functional axes, we predict, and then verify, that their weight-update directions are near-orthogonal, without any explicit orthogonality constraint during training. Orthogonal updates should combine without interfering, so parallel composition is a direct sum. The two adapters are simply added at inference, with no joint training, yielding few-step, streaming audio-video whose image quality tracks the bidirectional teacher. Compared to the chained baselines, the composed model matches or beats them on most metrics, making parallel composition a practical approach. The resulting streaming system generates joint audio-video in real time, $\approx$26 fps at $480\times832$ without quantization, and sustains 30 s of continuous generation with stable image quality.

Explore related subjects

cs.CV

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Jingyu Li, Xiaoxiao Xiang, Yiwen Guo. 2026-10-07. Real-Time Joint Audio-Video Generation by Parallel Adapter Composition. https://arxiv.org/abs/2610.10343

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

BiPO: Bidirectional Partial Occlusion Network for Text-to-Motion Synthesis

Generating natural and expressive human motions from textual descriptions is challenging due to the complexity of coordinating full-body dynamics and capturing nuanced motion patterns over extended sequences that accurately reflect the given text. To address this, we introduce BiPO, Bidirectional Partial Occlusion Network for Text-to-Motion Synthesis, a novel model that enhances text-to-motion synthesis by integrating part-based generation with a bidirectional autoregressive architecture. This integration allows BiPO to consider both past and future contexts during generation while enhancing detailed control over individual body parts without requiring ground-truth motion length. To relax the interdependency among body parts caused by the integration, we devise the Partial Occlusion technique, which probabilistically occludes the certain motion part information during training. In our comprehensive experiments, BiPO achieves state-of-the-art performance on the HumanML3D dataset, outperforming recent methods such as ParCo, MoMask, and BAMM in terms of FID scores and overall motion quality. Notably, BiPO excels not only in the text-to-motion generation task but also in motion editing tasks that synthesize motion based on partially generated motion sequences and textual descriptions. These results reveal the BiPO's effectiveness in advancing text-to-motion synthesis and its potential for practical applications.

cs.CV↗
arXiv2026

Batch Augmentation with Unimodal Fine-tuning for Multimodal Fusion of Large Language Models

In this paper, we propose batch augmentation with unimodal fine-tuning for multimodal learning. We start with pre-trained unimodal models. We fine-tune the unimodal models with the application data. After that, we form a Multi-Layer Perceptron (MLP) head that takes information from unimodal models and provides output. Finally, we train the MLP layer and unimodal parts with batch augmentation. Depending on the data, some unimodal models can be replaced by hard-coded scripts or AI agents. The unimodal training can also follow batch augmentation when the data is augmentable. We write a multimodal batch augmentation dataloader script that implements the batch augmentation for the multimodal data. We investigate the proposed method on the FPU23 ultrasound and UPMC Food-101 multimodal datasets. The multimodal large language model (LLM) with the proposed training achieves the best average result among the investigated methods across both datasets. According to our literature search, the proposed method achieves state-of-the-art (SOTA) accuracy of 93.29% on the UPMC Food-101 dataset, while we apply the ViT-L/16 model for vision and the GPT-2 model for text. We share the scripts of the proposed method with traditional counterparts at the following repository: github.com/dipuk0506/multimodal

cs.CV↗
arXiv2026

Geometry-Centered 3D Latent World Models for Growing Surfaces

Many physical systems do not merely move or deform; they grow, adding material and changing the geometry that a world model must represent. Existing world models are typically optimized for pixel prediction, reward prediction, or fixed-support physical dynamics, leaving open how to model systems whose underlying physical support expands over time and whose future morphology depends on hidden material response. We introduce FOLIAGE, a geometry-centered latent world model for growing surfaces. Within a fixed state budget, FOLIAGE represents mature regions as a compact scaffold while allocating higher-resolution state to regions predicted to drive near-future growth. This focuses representation and computation where new material and geometric change occur while retaining compact global context. FOLIAGE further separates observation, action, and privileged physics: heterogeneous RGB, point-cloud, and mesh observations are fused into a deployable geometric state; material controls condition the latent dynamics; and hidden physical energies guide training but are not required at deployment. To evaluate this setting, we introduce SURF-GARDEN and SURF-BENCH, providing controlled counterfactual branches, dense cross-modal correspondences, hidden physical signals, and stress tests for growing-geometry state learning. FOLIAGE reduces inverse-material error by $\approx40\%$ and 5-step mesh forecasting Chamfer error by $\approx30\%$ relative to strong baselines, while improving cross-modal retrieval by +14 mAP points. Stress tests show graceful degradation under sensor loss and correspondence corruption. On temporal 3D plant scans, FOLIAGE also improves passive future-geometry forecasting, while transfer experiments show that the learned geometry-centered state remains useful beyond the simulator.

cs.CV↗