Seed-ASR: Understanding Diverse Speech and Contexts with LLM-based Speech Recognition

Modern automatic speech recognition (ASR) model is required to accurately transcribe diverse speech signals (from different domains, languages, accents, etc) given the specific contextual information in various application scenarios. Classic end-to-end models fused with extra language models perform well, but mainly in data matching scenarios and are gradually approaching a bottleneck. In this work, we introduce Seed-ASR, a large language model (LLM) based speech recognition model. Seed-ASR is developed based on the framework of audio conditioned LLM (AcLLM), leveraging the capabilities of LLMs by inputting continuous speech representations together with contextual information into the LLM. Through stage-wise large-scale training and the elicitation of context-aware capabilities in LLM, Seed-ASR demonstrates significant improvement over end-to-end models on comprehensive evaluation sets, including multiple domains, accents/dialects and languages. Additionally, Seed-ASR can be further deployed to support specific needs in various scenarios without requiring extra language models. Compared to recently released large ASR models, Seed-ASR achieves 10%-40% reduction in word (or character, for Chinese) error rates on Chinese and English public test sets, further demonstrating its powerful performance.

Improving Audio Generation with Visual Enhanced Caption

Generative models have shown significant achievements in audio generation tasks. However, existing models struggle with complex and detailed prompts, leading to potential performance degradation. We hypothesize that this problem stems from the low quality and relatively small quantity of training data. In this work, we aim to create a large-scale audio dataset with rich captions for improving audio generation models. We develop an automated pipeline to generate detailed captions for audio-visual datasets by transforming predicted visual captions, audio captions, and tagging labels into comprehensive descriptions using a Large Language Model (LLM). We introduce Sound-VECaps, a dataset comprising 1.66M high-quality audio-caption pairs with enriched details including audio event orders, occurred places and environment information. We demonstrate that training with Sound-VECaps significantly enhances the capability of text-to-audio generation models to comprehend and generate audio from complex input prompts, improving overall system performance. Furthermore, we conduct ablation studies of Sound-VECaps across several audio-language tasks, suggesting its potential in advancing audio-text representation learning. Our dataset and models are available online.

Seed-TTS: A Family of High-Quality Versatile Speech Generation Models

We introduce Seed-TTS, a family of large-scale autoregressive text-to-speech (TTS) models capable of generating speech that is virtually indistinguishable from human speech. Seed-TTS serves as a foundation model for speech generation and excels in speech in-context learning, achieving performance in speaker similarity and naturalness that matches ground truth human speech in both objective and subjective evaluations. With fine-tuning, we achieve even higher subjective scores across these metrics. Seed-TTS offers superior controllability over various speech attributes such as emotion and is capable of generating highly expressive and diverse speech for speakers in the wild. Furthermore, we propose a self-distillation method for speech factorization, as well as a reinforcement learning approach to enhance model robustness, speaker similarity, and controllability. We additionally present a non-autoregressive (NAR) variant of the Seed-TTS model, named $\text{Seed-TTS}_\text{DiT}$, which utilizes a fully diffusion-based architecture. Unlike previous NAR-based TTS systems, $\text{Seed-TTS}_\text{DiT}$ does not depend on pre-estimated phoneme durations and performs speech generation through end-to-end processing. We demonstrate that this variant achieves comparable performance to the language model-based variant and showcase its effectiveness in speech editing. We encourage readers to listen to demos at \url{https://bytedancespeech.github.io/seedtts_tech_report}.

Open Challenges and Opportunities in Federated Foundation Models Towards Biomedical Healthcare

This survey explores the transformative impact of foundation models (FMs) in artificial intelligence, focusing on their integration with federated learning (FL) for advancing biomedical research. Foundation models such as ChatGPT, LLaMa, and CLIP, which are trained on vast datasets through methods including unsupervised pretraining, self-supervised learning, instructed fine-tuning, and reinforcement learning from human feedback, represent significant advancements in machine learning. These models, with their ability to generate coherent text and realistic images, are crucial for biomedical applications that require processing diverse data forms such as clinical reports, diagnostic images, and multimodal patient interactions. The incorporation of FL with these sophisticated models presents a promising strategy to harness their analytical power while safeguarding the privacy of sensitive medical data. This approach not only enhances the capabilities of FMs in medical diagnostics and personalized treatment but also addresses critical concerns about data privacy and security in healthcare. This survey reviews the current applications of FMs in federated settings, underscores the challenges, and identifies future research directions including scaling FMs, managing data diversity, and enhancing communication efficiency within FL frameworks. The objective is to encourage further research into the combined potential of FMs and FL, laying the groundwork for groundbreaking healthcare innovations.

VoiceShop: A Unified Speech-to-Speech Framework for Identity-Preserving Zero-Shot Voice Editing

We present VoiceShop, a novel speech-to-speech framework that can modify multiple attributes of speech, such as age, gender, accent, and speech style, in a single forward pass while preserving the input speaker's timbre. Previous works have been constrained to specialized models that can only edit these attributes individually and suffer from the following pitfalls: the magnitude of the conversion effect is weak, there is no zero-shot capability for out-of-distribution speakers, or the synthesized outputs exhibit undesirable timbre leakage. Our work proposes solutions for each of these issues in a simple modular framework based on a conditional diffusion backbone model with optional normalizing flow-based and sequence-to-sequence speaker attribute-editing modules, whose components can be combined or removed during inference to meet a wide array of tasks without additional model finetuning. Audio samples are available at \url{https://voiceshopai.github.io}.

CMP: Cooperative Motion Prediction with Multi-Agent Communication

The confluence of the advancement of Autonomous Vehicles (AVs) and the maturity of Vehicle-to-Everything (V2X) communication has enabled the capability of cooperative connected and automated vehicles (CAVs). Building on top of cooperative perception, this paper explores the feasibility and effectiveness of cooperative motion prediction. Our method, CMP, takes LiDAR signals as input to enhance tracking and prediction capabilities. Unlike previous work that focuses separately on either cooperative perception or motion prediction, our framework, to the best of our knowledge, is the first to address the unified problem where CAVs share information in both perception and prediction modules. Incorporated into our design is the unique capability to tolerate realistic V2X bandwidth limitations and transmission delays, while dealing with bulky perception representations. We also propose a prediction aggregation module, which unifies the predictions obtained by different CAVs and generates the final prediction. Through extensive experiments and ablation studies, we demonstrate the effectiveness of our method in cooperative perception, tracking, and motion prediction tasks. In particular, CMP reduces the average prediction error by 17.2\% with fewer missing detections compared with the no cooperation setting. Our work marks a significant step forward in the cooperative capabilities of CAVs, showcasing enhanced performance in complex scenarios.

StreamVoice: Streamable Context-Aware Language Modeling for Real-time Zero-Shot Voice Conversion

Recent language model (LM) advancements have showcased impressive zero-shot voice conversion (VC) performance. However, existing LM-based VC models usually apply offline conversion from source semantics to acoustic features, demanding the complete source speech, and limiting their deployment to real-time applications. In this paper, we introduce StreamVoice, a novel streaming LM-based model for zero-shot VC, facilitating real-time conversion given arbitrary speaker prompts and source speech. Specifically, to enable streaming capability, StreamVoice employs a fully causal context-aware LM with a temporal-independent acoustic predictor, while alternately processing semantic and acoustic features at each time step of autoregression which eliminates the dependence on complete source speech. To address the potential performance degradation from the incomplete context in streaming processing, we enhance the context-awareness of the LM through two strategies: 1) teacher-guided context foresight, using a teacher model to summarize the present and future semantic context during training to guide the model's forecasting for missing context; 2) semantic masking strategy, promoting acoustic prediction from preceding corrupted semantic and acoustic input, enhancing context-learning ability. Notably, StreamVoice is the first LM-based streaming zero-shot VC model without any future look-ahead. Experimental results demonstrate StreamVoice's streaming conversion capability while maintaining zero-shot performance comparable to non-streaming VC systems.

Towards the Unification of Generative and Discriminative Visual Foundation Model: A Survey

The advent of foundation models, which are pre-trained on vast datasets, has ushered in a new era of computer vision, characterized by their robustness and remarkable zero-shot generalization capabilities. Mirroring the transformative impact of foundation models like large language models (LLMs) in natural language processing, visual foundation models (VFMs) have become a catalyst for groundbreaking developments in computer vision. This review paper delineates the pivotal trajectories of VFMs, emphasizing their scalability and proficiency in generative tasks such as text-to-image synthesis, as well as their adeptness in discriminative tasks including image segmentation. While generative and discriminative models have historically charted distinct paths, we undertake a comprehensive examination of the recent strides made by VFMs in both domains, elucidating their origins, seminal breakthroughs, and pivotal methodologies. Additionally, we collate and discuss the extensive resources that facilitate the development of VFMs and address the challenges that pave the way for future research endeavors. A crucial direction for forthcoming innovation is the amalgamation of generative and discriminative paradigms. The nascent application of generative models within discriminative contexts signifies the early stages of this confluence. This survey aspires to be a contemporary compendium for scholars and practitioners alike, charting the course of VFMs and illuminating their multifaceted landscape.

Novel View Synthesis from a Single RGBD Image for Indoor Scenes

In this paper, we propose an approach for synthesizing novel view images from a single RGBD (Red Green Blue-Depth) input. Novel view synthesis (NVS) is an interesting computer vision task with extensive applications. Methods using multiple images has been well-studied, exemplary ones include training scene-specific Neural Radiance Fields (NeRF), or leveraging multi-view stereo (MVS) and 3D rendering pipelines. However, both are either computationally intensive or non-generalizable across different scenes, limiting their practical value. Conversely, the depth information embedded in RGBD images unlocks 3D potential from a singular view, simplifying NVS. The widespread availability of compact, affordable stereo cameras, and even LiDARs in contemporary devices like smartphones, makes capturing RGBD images more accessible than ever. In our method, we convert an RGBD image into a point cloud and render it from a different viewpoint, then formulate the NVS task into an image translation problem. We leveraged generative adversarial networks to style-transfer the rendered image, achieving a result similar to a photograph taken from the new perspective. We explore both unsupervised learning using CycleGAN and supervised learning with Pix2Pix, and demonstrate the qualitative results. Our method circumvents the limitations of traditional multi-image techniques, holding significant promise for practical, real-time applications in NVS.

EqDrive: Efficient Equivariant Motion Forecasting with Multi-Modality for Autonomous Driving

Forecasting vehicular motions in autonomous driving requires a deep understanding of agent interactions and the preservation of motion equivariance under Euclidean geometric transformations. Traditional models often lack the sophistication needed to handle the intricate dynamics inherent to autonomous vehicles and the interaction relationships among agents in the scene. As a result, these models have a lower model capacity, which then leads to higher prediction errors and lower training efficiency. In our research, we employ EqMotion, a leading equivariant particle, and human prediction model that also accounts for invariant agent interactions, for the task of multi-agent vehicle motion forecasting. In addition, we use a multi-modal prediction mechanism to account for multiple possible future paths in a probabilistic manner. By leveraging EqMotion, our model achieves state-of-the-art (SOTA) performance with fewer parameters (1.2 million) and a significantly reduced training time (less than 2 hours).