Search Results for author:
Found 12 papers, 0 papers with code
The Sogou-TIIC Speech Translation System for IWSLT 2018
Our final average result on speech translation is 31. 02 BLEU.
High-dimensional Asymptotics of Feature Learning: How One Gradient Step Improves the Representation
We study the first gradient descent step on the first-layer parameters $\boldsymbol{W}$ in a two-layer neural network: $f(\boldsymbol{x}) = \frac{1}{\sqrt{N}}\boldsymbol{a}^\top\sigma(\boldsymbol{W}^\top\boldsymbol{x})$, where $\boldsymbol{W}\in\mathbb{R}^{d\times N}, \boldsymbol{a}\in\mathbb{R}^{N}$ are randomly initialized, and the training objective is the empirical MSE loss: $\frac{1}{n}\sum_{i=1}^n (f(\boldsymbol{x}_i)-y_i)^2$.
Causal Disentanglement for Semantics-Aware Intent Learning in Recommendation
Traditional recommendation models trained on observational interaction data have generated large impacts in a wide range of applications, it faces bias problems that cover users' true intent and thus deteriorate the recommendation effectiveness.
IQDUBBING: Prosody modeling based on discrete self-supervised speech representation for expressive voice conversion
Specifically, prosody vector is first extracted from pre-trained VQ-Wav2Vec model, where rich prosody information is embedded while most speaker and environment information are removed effectively by quantization.
Multi-speaker Multi-style Text-to-speech Synthesis With Single-speaker Single-style Training Data Scenarios
Moreover, the explicit prosody features used in the prosody predicting module can increase the diversity of synthetic speech by adjusting the value of prosody features.
One-shot Voice Conversion For Style Transfer Based On Speaker Adaptation
One-shot style transfer is a challenging task, since training on one utterance makes model extremely easy to over-fit to training data and causes low speaker similarity and lack of expressiveness.
Deformed semicircle law and concentration of nonlinear random matrices for ultra-wide neural networks
In this paper, we study the two-layer fully connected neural network given by $f(X)=\frac{1}{\sqrt{d_1}}\boldsymbol{a}^\top\sigma\left(WX\right)$, where $X\in\mathbb{R}^{d_0\times n}$ is a deterministic data matrix, $W\in\mathbb{R}^{d_1\times d_0}$ and $\boldsymbol{a}\in\mathbb{R}^{d_1}$ are random Gaussian weights, and $\sigma$ is a nonlinear activation function.
Hilbert Sinkhorn Divergence for Optimal Transport
Sinkhorn divergence has become a very popular metric to compare probability distributions in optimal transport.
Enriching Source Style Transfer in Recognition-Synthesis based Non-Parallel Voice Conversion
Specifically, prosodic features are used to explicit model prosody, while VAE and reference encoder are used to implicitly model prosody, which take Mel spectrum and bottleneck feature as input respectively.
WNARS: WFST based Non-autoregressive Streaming End-to-End Speech Recognition
Recently, attention-based encoder-decoder (AED) end-to-end (E2E) models have drawn more and more attention in the field of automatic speech recognition (ASR).
Spectra of the Conjugate Kernel and Neural Tangent Kernel for linear-width neural networks
We study the eigenvalue distributions of the Conjugate Kernel and Neural Tangent Kernel associated to multi-layer feedforward neural networks.
Polynomial Representation for Persistence Diagram
In this work, we discover a set of general polynomials that vanish on vectorized PDs and extract the task-adapted feature representation from these polynomials.
