SANA-Streaming introduces a novel, efficient framework that enables real-time, high-resolution streaming video-to-video editing by combining a hybrid diffusion transformer with specialized training an…

This paper introduces a training recipe for sentence-level and long-form streaming speech-to-speech translation using only 2k hours of paired cross-lingual data and auxiliary supervision.

The paper introduces X-Stream, a new benchmark for multi-stream video understanding, and finds that current state-of-the-art MLLMs perform poorly when required to process multiple concurrent video str…

LongLive-RAG proposes a novel Retrieval-Augmented Generation (RAG) framework to stabilize and improve the quality of long-horizon video generation by treating the entire generated history as a searcha…

Murmur is an efficient inference system for long-form ASR that resolves the accuracy-latency trade-off by optimizing both inter-chunk processing and intra-chunk attention mechanisms.

This paper proposes LongE2V, a method for high-quality video recovery from sparse event streams using pre-trained video diffusion priors.

This paper introduces WanSong, a diffusion-based model for long-form, commercial-grade song generation that directly generates high-fidelity, multilingual songs up to 5 minutes and outputs dual stems,…

This paper proposes a framework for federated learning with limited client memory and time-varying sampling costs, and develops an Active-Constraint Drift-Plus-Penalty (ACDPP) policy to minimize the c…

The paper introduces an efficient, novel algorithm for incremental Byte Pair Encoding (BPE) tokenization that processes input text prefix by prefix, achieving significant speedups and enabling streami…

This paper proposes cumsum-composable phase transport, a streaming-native temporal layer for keyword spotting using unitary transport, prefix differences, and gated residual updates.

The paper introduces YouTube-Synch, a robust system that successfully replicates content from thousands of YouTube channels to decentralized storage by continuously adapting to and bypassing YouTube's…

This paper proves that no single-pass semi-streaming algorithm can achieve a better-than-half approximation to the maximum matching problem, implying the optimality of the naive greedy algorithm.

SHARP proposes a novel sleep-based hierarchical replay framework to efficiently learn long-range non-stationary temporal patterns in streaming data, achieving improved context retention and predictive…

The paper develops a new framework for proving lower bounds for the maximum matching problem in the semi-streaming model, improving upon the previous best known bounds.

JenBridge is a novel, adaptive framework that generates high-fidelity, long-form video soundtracks, significantly improving narrative coherence and naturalness across scene transitions.

This paper proposes a framework for creating low-latency, interactive generative music AI using distillation in a streaming autoregressive latent space and music-aware consistency objectives.