While many recent any-to-any voice conversion models succeed in transferring some target speech's style information to the converted speech, they still lack the ability to faithfully reproduce the speaking style of the target speaker. In this work, we propose a novel method to extract rich style information from target utterances and to efficiently transfer it to source speech content without requiring text transcriptions or speaker labeling. Our proposed approach introduces an attention mechanism utilizing a self-supervised learning (SSL) model to collect the speaking styles of a target speaker each corresponding to the different phonetic content. The styles are represented with a set of embeddings called stylebook. In the next step, the stylebook is attended with the source speech's phonetic content to determine the final target style for each source content. Finally, content information extracted from the source speech and content-dependent target style embeddings are fed into a diffusion-based decoder to generate the converted speech mel-spectrogram. Experiment results show that our proposed method combined with a diffusion-based generative model can achieve better speaker similarity in any-to-any voice conversion tasks when compared to baseline models, while the increase in computational complexity with longer utterances is suppressed.
相關內容
Despite outstanding performance in many tasks, language models are notoriously inclined to make factual errors in tasks requiring arithmetic computation. We address this deficiency by creating Calc-X, a collection of datasets that demonstrates the appropriate use of a calculator in reasoning chains. Calc-X is suitable for teaching language models to offload computations to a symbolic system. We survey and unify several existing chain-of-thought datasets into a proposed format, resulting in a standard collection of over 300,000 samples requiring arithmetic reasoning. Finally, we use the new Calc-X collection to train open-source calculator-using models we call Calcformers and show that these models approximately double the accuracy of generating correct results compared to vanilla language model baselines. We make all Calc-X datasets, source code and Calcformers models publicly available.
Neural document rerankers are extremely effective in terms of accuracy. However, the best models require dedicated hardware for serving, which is costly and often not feasible. To avoid this serving-time requirement, we present a method of capturing up to 86% of the gains of a Transformer cross-attention model with a lexicalized scoring function that only requires 10-6% of the Transformer's FLOPs per document and can be served using commodity CPUs. When combined with a BM25 retriever, this approach matches the quality of a state-of-the art dual encoder retriever, that still requires an accelerator for query encoding. We introduce NAIL (Non-Autoregressive Indexing with Language models) as a model architecture that is compatible with recent encoder-decoder and decoder-only large language models, such as T5, GPT-3 and PaLM. This model architecture can leverage existing pre-trained checkpoints and can be fine-tuned for efficiently constructing document representations that do not require neural processing of queries.
To alleviate the heavy annotation burden for training a reliable crowd counting model and thus make the model more practicable and accurate by being able to benefit from more data, this paper presents a new semi-supervised method based on the mean teacher framework. When there is a scarcity of labeled data available, the model is prone to overfit local patches. Within such contexts, the conventional approach of solely improving the accuracy of local patch predictions through unlabeled data proves inadequate. Consequently, we propose a more nuanced approach: fostering the model's intrinsic 'subitizing' capability. This ability allows the model to accurately estimate the count in regions by leveraging its understanding of the crowd scenes, mirroring the human cognitive process. To achieve this goal, we apply masking on unlabeled data, guiding the model to make predictions for these masked patches based on the holistic cues. Furthermore, to help with feature learning, herein we incorporate a fine-grained density classification task. Our method is general and applicable to most existing crowd counting methods as it doesn't have strict structural or loss constraints. In addition, we observe that the model trained with our framework exhibits a 'subitizing'-like behavior. It accurately predicts low-density regions with only a 'glance', while incorporating local details to predict high-density regions. Our method achieves the state-of-the-art performance, surpassing previous approaches by a large margin on challenging benchmarks such as ShanghaiTech A and UCF-QNRF. The code is available at: //github.com/cha15yq/MRC-Crowd.
Retrieval-augmented language models show promise in addressing issues like outdated information and hallucinations in language models (LMs). However, current research faces two main problems: 1) determining what information to retrieve, and 2) effectively combining retrieved information during generation. We argue that valuable retrieved information should not only be related to the current source text but also consider the future target text, given the nature of LMs that model future tokens. Moreover, we propose that aggregation using latent variables derived from a compact latent space is more efficient than utilizing explicit raw text, which is limited by context length and susceptible to noise. Therefore, we introduce RegaVAE, a retrieval-augmented language model built upon the variational auto-encoder (VAE). It encodes the text corpus into a latent space, capturing current and future information from both source and target text. Additionally, we leverage the VAE to initialize the latent space and adopt the probabilistic form of the retrieval generation paradigm by expanding the Gaussian prior distribution into a Gaussian mixture distribution. Theoretical analysis provides an optimizable upper bound for RegaVAE. Experimental results on various datasets demonstrate significant improvements in text generation quality and hallucination removal.
This Paper proposes a novel Transformer-based end-to-end autonomous driving model named Detrive. This model solves the problem that the past end-to-end models cannot detect the position and size of traffic participants. Detrive uses an end-to-end transformer based detection model as its perception module; a multi-layer perceptron as its feature fusion network; a recurrent neural network with gate recurrent unit for path planning; and two controllers for the vehicle's forward speed and turning angle. The model is trained with an on-line imitation learning method. In order to obtain a better training set, a reinforcement learning agent that can directly obtain a ground truth bird's-eye view map from the Carla simulator as a perceptual output, is used as teacher for the imitation learning. The trained model is tested on the Carla's autonomous driving benchmark. The results show that the Transformer detector based end-to-end model has obvious advantages in dynamic obstacle avoidance compared with the traditional classifier based end-to-end model.
In task-oriented dialogue scenarios, cross-domain zero-shot slot filling plays a vital role in leveraging source domain knowledge to learn a model with high generalization ability in unknown target domain where annotated data is unavailable. However, the existing state-of-the-art zero-shot slot filling methods have limited generalization ability in target domain, they only show effective knowledge transfer on seen slots and perform poorly on unseen slots. To alleviate this issue, we present a novel Hierarchical Contrastive Learning Framework (HiCL) for zero-shot slot filling. Specifically, we propose a coarse- to fine-grained contrastive learning based on Gaussian-distributed embedding to learn the generalized deep semantic relations between utterance-tokens, by optimizing inter- and intra-token distribution distance. This encourages HiCL to generalize to the slot types unseen at training phase. Furthermore, we present a new iterative label set semantics inference method to unbiasedly and separately evaluate the performance of unseen slot types which entangled with their counterparts (i.e., seen slot types) in the previous zero-shot slot filling evaluation methods. The extensive empirical experiments on four datasets demonstrate that the proposed method achieves comparable or even better performance than the current state-of-the-art zero-shot slot filling approaches.
Restricted mean survival time (RMST) models have gained popularity when analyzing time-to-event outcomes because RMST models offer more straightforward interpretations of treatment effects with fewer assumptions than hazard ratios commonly estimated from Cox models. However, few network meta-analysis (NMA) methods have been developed using RMST. In this paper, we propose advanced RMST NMA models when individual participant data are available. Our models allow us to study treatment effect moderation and provide comprehensive understanding about comparative effectiveness of treatments and subgroup effects. An extensive simulation study and a real data example about treatments for patients with atrial fibrillation are presented.
Widely used language models (LMs) are typically built by scaling up a two-stage training pipeline: a pre-training stage that uses a very large, diverse dataset of text and a fine-tuning (sometimes, 'alignment') stage that uses targeted examples or other specifications of desired behaviors. While it has been hypothesized that knowledge and skills come from pre-training, and fine-tuning mostly filters this knowledge and skillset, this intuition has not been extensively tested. To aid in doing so, we introduce a novel technique for decoupling the knowledge and skills gained in these two stages, enabling a direct answer to the question, "What would happen if we combined the knowledge learned by a large model during pre-training with the knowledge learned by a small model during fine-tuning (or vice versa)?" Using an RL-based framework derived from recent developments in learning from human preferences, we introduce emulated fine-tuning (EFT), a principled and practical method for sampling from a distribution that approximates (or 'emulates') the result of pre-training and fine-tuning at different scales. Our experiments with EFT show that scaling up fine-tuning tends to improve helpfulness, while scaling up pre-training tends to improve factuality. Beyond decoupling scale, we show that EFT enables test-time adjustment of competing behavioral traits like helpfulness and harmlessness without additional training. Finally, a special case of emulated fine-tuning, which we call LM up-scaling, avoids resource-intensive fine-tuning of large pre-trained models by ensembling them with small fine-tuned models, essentially emulating the result of fine-tuning the large pre-trained model. Up-scaling consistently improves helpfulness and factuality of instruction-following models in the Llama, Llama-2, and Falcon families, without additional hyperparameters or training.
Generative models are now capable of producing highly realistic images that look nearly indistinguishable from the data on which they are trained. This raises the question: if we have good enough generative models, do we still need datasets? We investigate this question in the setting of learning general-purpose visual representations from a black-box generative model rather than directly from data. Given an off-the-shelf image generator without any access to its training data, we train representations from the samples output by this generator. We compare several representation learning methods that can be applied to this setting, using the latent space of the generator to generate multiple "views" of the same semantic content. We show that for contrastive methods, this multiview data can naturally be used to identify positive pairs (nearby in latent space) and negative pairs (far apart in latent space). We find that the resulting representations rival those learned directly from real data, but that good performance requires care in the sampling strategy applied and the training method. Generative models can be viewed as a compressed and organized copy of a dataset, and we envision a future where more and more "model zoos" proliferate while datasets become increasingly unwieldy, missing, or private. This paper suggests several techniques for dealing with visual representation learning in such a future. Code is released on our project page: //ali-design.github.io/GenRep/
We propose to pre-train a unified language model for both autoencoding and partially autoregressive language modeling tasks using a novel training procedure, referred to as a pseudo-masked language model (PMLM). Given an input text with masked tokens, we rely on conventional masks to learn inter-relations between corrupted tokens and context via autoencoding, and pseudo masks to learn intra-relations between masked spans via partially autoregressive modeling. With well-designed position embeddings and self-attention masks, the context encodings are reused to avoid redundant computation. Moreover, conventional masks used for autoencoding provide global masking information, so that all the position embeddings are accessible in partially autoregressive language modeling. In addition, the two tasks pre-train a unified language model as a bidirectional encoder and a sequence-to-sequence decoder, respectively. Our experiments show that the unified language models pre-trained using PMLM achieve new state-of-the-art results on a wide range of natural language understanding and generation tasks across several widely used benchmarks.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191