arXiv subjects

Yi Ren

Publications and source records attributed to Yi Ren.

At least 163 records · Page 9Linked to original sources

arXiv2023

Item Cold Start Recommendation via Adversarial Variational Auto-encoder Warm-up

The gap between the randomly initialized item ID embedding and the well-trained warm item ID embedding makes the cold items hard to suit the recommendation system, which is trained on the data of historical warm items. To alleviate the performance decline of new items recommendation, the distribution of the new item ID embedding should be close to that of the historical warm items. To achieve this goal, we propose an Adversarial Variational Auto-encoder Warm-up model (AVAEW) to generate warm-up item ID embedding for cold items. Specifically, we develop a conditional variational auto-encoder model to leverage the side information of items for generating the warm-up item ID embedding. Particularly, we introduce an adversarial module to enforce the alignment between warm-up item ID embedding distribution and historical item ID embedding distribution. We demonstrate the effectiveness and compatibility of the proposed method by extensive offline experiments on public datasets and online A/B tests on a real-world large-scale news recommendation platform.

cs.IR↗

arXiv2023

Approximating Discontinuous Nash Equilibrial Values of Two-Player General-Sum Differential Games

Finding Nash equilibrial policies for two-player differential games requires solving Hamilton-Jacobi-Isaacs (HJI) PDEs. Self-supervised learning has been used to approximate solutions of such PDEs while circumventing the curse of dimensionality. However, this method fails to learn discontinuous PDE solutions due to its sampling nature, leading to poor safety performance of the resulting controllers in robotics applications when player rewards are discontinuous. This paper investigates two potential solutions to this problem: a hybrid method that leverages both supervised Nash equilibria and the HJI PDE, and a value-hardening method where a sequence of HJIs are solved with a gradually hardening reward. We compare these solutions using the resulting generalization and safety performance in two vehicle interaction simulation studies with 5D and 9D state spaces, respectively. Results show that with informative supervision (e.g., collision and near-collision demonstrations) and the low cost of self-supervised learning, the hybrid method achieves better safety performance than the supervised, self-supervised, and value hardening approaches on equal computational budget. Value hardening fails to generalize in the higher-dimensional case without informative supervision. Lastly, we show that the neural activation function needs to be continuously differentiable for learning PDEs and its choice can be case dependent.

cs.LG↗

arXiv2023

Slate-Aware Ranking for Recommendation

We see widespread adoption of slate recommender systems, where an ordered item list is fed to the user based on the user interests and items' content. For each recommendation, the user can select one or several items from the list for further interaction. In this setting, the significant impact on user behaviors from the mutual influence among the items is well understood. The existing methods add another step of slate re-ranking after the ranking stage of recommender systems, which considers the mutual influence among recommended items to re-rank and generate the recommendation results so as to maximize the expected overall utility. However, to model the complex interaction of multiple recommended items, the re-ranking stage usually can just handle dozens of candidates because of the constraint of limited hardware resource and system latency. Therefore, the ranking stage is still essential for most applications to provide high-quality candidate set for the re-ranking stage. In this paper, we propose a solution named Slate-Aware ranking (SAR) for the ranking stage. By implicitly considering the relations among the slate items, it significantly enhances the quality of the re-ranking stage's candidate set and boosts the relevance and diversity of the overall recommender systems. Both experiments with the public datasets and internal online A/B testing are conducted to verify its effectiveness.

cs.IR↗

arXiv2023

How to prepare your task head for finetuning

In deep learning, transferring information from a pretrained network to a downstream task by finetuning has many benefits. The choice of task head plays an important role in fine-tuning, as the pretrained and downstream tasks are usually different. Although there exist many different designs for finetuning, a full understanding of when and why these algorithms work has been elusive. We analyze how the choice of task head controls feature adaptation and hence influences the downstream performance. By decomposing the learning dynamics of adaptation, we find that the key aspect is the training accuracy and loss at the beginning of finetuning, which determines the "energy" available for the feature's adaptation. We identify a significant trend in the effect of changes in this initial energy on the resulting features after fine-tuning. Specifically, as the energy increases, the Euclidean and cosine distances between the resulting and original features increase, while their dot products (and the resulting features' norm) first increase and then decrease. Inspired by this, we give several practical principles that lead to better downstream performance. We analytically prove this trend in an overparamterized linear setting and verify its applicability to different experimental settings.

cs.LG↗

arXiv2023

GeneFace: Generalized and High-Fidelity Audio-Driven 3D Talking Face Synthesis

Generating photo-realistic video portrait with arbitrary speech audio is a crucial problem in film-making and virtual reality. Recently, several works explore the usage of neural radiance field in this task to improve 3D realness and image fidelity. However, the generalizability of previous NeRF-based methods to out-of-domain audio is limited by the small scale of training data. In this work, we propose GeneFace, a generalized and high-fidelity NeRF-based talking face generation method, which can generate natural results corresponding to various out-of-domain audio. Specifically, we learn a variaitional motion generator on a large lip-reading corpus, and introduce a domain adaptative post-net to calibrate the result. Moreover, we learn a NeRF-based renderer conditioned on the predicted facial motion. A head-aware torso-NeRF is proposed to eliminate the head-torso separation problem. Extensive experiments show that our method achieves more generalized and high-fidelity talking face generation compared to previous methods.

cs.CV↗

arXiv2023

Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Models

Large-scale multimodal generative modeling has created milestones in text-to-image and text-to-video generation. Its application to audio still lags behind for two main reasons: the lack of large-scale datasets with high-quality text-audio pairs, and the complexity of modeling long continuous audio data. In this work, we propose Make-An-Audio with a prompt-enhanced diffusion model that addresses these gaps by 1) introducing pseudo prompt enhancement with a distill-then-reprogram approach, it alleviates data scarcity with orders of magnitude concept compositions by using language-free audios; 2) leveraging spectrogram autoencoder to predict the self-supervised audio representation instead of waveforms. Together with robust contrastive language-audio pretraining (CLAP) representations, Make-An-Audio achieves state-of-the-art results in both objective and subjective benchmark evaluation. Moreover, we present its controllability and generalization for X-to-Audio with "No Modality Left Behind", for the first time unlocking the ability to generate high-definition, high-fidelity audios given a user-defined modality input. Audio samples are available at https://Text-to-Audio.github.io

cs.SD↗

arXiv2023

XNLI: Explaining and Diagnosing NLI-based Visual Data Analysis

Natural language interfaces (NLIs) enable users to flexibly specify analytical intentions in data visualization. However, diagnosing the visualization results without understanding the underlying generation process is challenging. Our research explores how to provide explanations for NLIs to help users locate the problems and further revise the queries. We present XNLI, an explainable NLI system for visual data analysis. The system introduces a Provenance Generator to reveal the detailed process of visual transformations, a suite of interactive widgets to support error adjustments, and a Hint Generator to provide query revision hints based on the analysis of user queries and interactions. Two usage scenarios of XNLI and a user study verify the effectiveness and usability of the system. Results suggest that XNLI can significantly enhance task accuracy without interrupting the NLI-based analysis process.

cs.HC↗

arXiv2022

Waveguide-Integrated Two-Dimensional Material Photodetectors in Thin-Film Lithium Niobate

Thin-film lithium niobate on insulator (LNOI) is a promising platform for optical communications, microwave photonics, and quantum technologies. While many high-performance devices like electro-optic modulators and frequency comb sources have been achieved on LNOI platform, it remains challenging to realize photodetectors (PDs) on LNOI platform using simple and low-cost fabrication techniques. Two-dimensional (2D) materials are excellent candidates to achieve photodetection since they feature strong light-matter interaction, excellent mechanical flexibility, and potential large-scale complementary metal-oxide-semiconductor-compatible fabrication. In this work, we propose to address this demand using an LNOI-2D material platform and demonstrate two types of high-performance LNOI waveguide-integrated 2D material PDs, namely graphene and Tellurium (Te). Specifically, the LNOI-graphene PD features broadband operations at telecom and visible wavelengths, high normalized photocurrent-to-dark current ratios up to 3*106 W-1, and large 3-dB photoelectric bandwidths of over 40 GHz, simultaneously. The LNOI-Te PD on the other hand provides an ultrahigh responsivity of 7 A/W under 0.5 V bias for telecom optical signals while supporting GHz frequency responses. Our results show that the versatile properties of 2D materials and their excellent compatibility with LNOI waveguides could provide important low-cost solutions for system operating point monitoring and high-speed photoelectric conversion in future LN photonic integrated circuits.

physics.optics↗

arXiv2022

Principles for generation of reverberation

In modern neuroscience, memory has been postulated to stored in neural circuits as sequential spike train and Reverberation is one of the specific example.Former research has made much progress on phenomenon description. However, the mechanism of reverberation has been unclear yet. In this study, combining electrophysiological record and numerical simulation, we confirmed a formerly unrealized neuron property that is necessary for the burst generation in reverberation. Secondly, we find out the mechanism of sequential pattern generation which clearly explained by network topology and asynchronous neurotransmitter release. In addition, we also developed a pipeline that could design the network fire in manually set order. Thirdly, we explored the dynamics of STDP learning and chased down the effects of STDP Rule in reverberation. With these understandings, we developed a STDP based learning rule which could drive the network to remember any presupposed sequence. These results indicated that neuron circuit can remember malformation through STDP rule. Those information are stored in synapse connections. By this way, animals remember information as spike sequence pattern.

q-bio.NC↗

arXiv2022

VarietySound: Timbre-Controllable Video to Sound Generation via Unsupervised Information Disentanglement

Video to sound generation aims to generate realistic and natural sound given a video input. However, previous video-to-sound generation methods can only generate a random or average timbre without any controls or specializations of the generated sound timbre, leading to the problem that people cannot obtain the desired timbre under these methods sometimes. In this paper, we pose the task of generating sound with a specific timbre given a video input and a reference audio sample. To solve this task, we disentangle each target sound audio into three components: temporal information, acoustic information, and background information. We first use three encoders to encode these components respectively: 1) a temporal encoder to encode temporal information, which is fed with video frames since the input video shares the same temporal information as the original audio; 2) an acoustic encoder to encode timbre information, which takes the original audio as input and discards its temporal information by a temporal-corrupting operation; and 3) a background encoder to encode the residual or background sound, which uses the background part of the original audio as input. To make the generated result achieve better quality and temporal alignment, we also adopt a mel discriminator and a temporal discriminator for the adversarial training. Our experimental results on the VAS dataset demonstrate that our method can generate high-quality audio samples with good synchronization with events in video and high timbre similarity with the reference audio.

cs.MM↗

arXiv2022

Safety-Critical Optimal Control for Robotic Manipulators in A Cluttered Environment

Designing safety-critical control for robotic manipulators is challenging, especially in a cluttered environment. First, the actual trajectory of a manipulator might deviate from the planned one due to the complex collision environments and non-trivial dynamics, leading to collision; Second, the feasible space for the manipulator is hard to obtain since the explicit distance functions between collision meshes are unknown. By analyzing the relationship between the safe set and the controlled invariant set, this paper proposes a data-driven control barrier function (CBF) construction method, which extracts CBF from distance samples. Specifically, the CBF guarantees the controlled invariant property for considering the system dynamics. The data-driven method samples the distance function and determines the safe set. Then, the CBF is synthesized based on the safe set by a scenario-based sum of square (SOS) program. Unlike most existing linearization based approaches, our method reserves the volume of the feasible space for planning without approximation, which helps find a solution in a cluttered environment. The control law is obtained by solving a CBF-based quadratic program in real time, which works as a safe filter for the desired planning-based controller. Moreover, our method guarantees safety with the proven probabilistic result. Our method is validated on a 7-DOF manipulator in both real and virtual cluttered environments. The experiments show that the manipulator is able to execute tasks where the clearance between obstacles is in millimeters.

cs.RO↗

arXiv2022

SDMuse: Stochastic Differential Music Editing and Generation via Hybrid Representation

While deep generative models have empowered music generation, it remains a challenging and under-explored problem to edit an existing musical piece at fine granularity. In this paper, we propose SDMuse, a unified Stochastic Differential Music editing and generation framework, which can not only compose a whole musical piece from scratch, but also modify existing musical pieces in many ways, such as combination, continuation, inpainting, and style transferring. The proposed SDMuse follows a two-stage pipeline to achieve music generation and editing on top of a hybrid representation including pianoroll and MIDI-event. In particular, SDMuse first generates/edits pianoroll by iteratively denoising through a stochastic differential equation (SDE) based on a diffusion model generative prior, and then refines the generated pianoroll and predicts MIDI-event tokens auto-regressively. We evaluate the generated music of our method on ailabs1k7 pop music dataset in terms of quality and controllability on various music editing and generation tasks. Experimental results demonstrate the effectiveness of our proposed stochastic differential music editing and generation process, as well as the hybrid representations.

cs.SD↗

arXiv2022

Pseudo Numerical Methods for Diffusion Models on Manifolds

Denoising Diffusion Probabilistic Models (DDPMs) can generate high-quality samples such as image and audio samples. However, DDPMs require hundreds to thousands of iterations to produce final samples. Several prior works have successfully accelerated DDPMs through adjusting the variance schedule (e.g., Improved Denoising Diffusion Probabilistic Models) or the denoising equation (e.g., Denoising Diffusion Implicit Models (DDIMs)). However, these acceleration methods cannot maintain the quality of samples and even introduce new noise at a high speedup rate, which limit their practicability. To accelerate the inference process while keeping the sample quality, we provide a fresh perspective that DDPMs should be treated as solving differential equations on manifolds. Under such a perspective, we propose pseudo numerical methods for diffusion models (PNDMs). Specifically, we figure out how to solve differential equations on manifolds and show that DDIMs are simple cases of pseudo numerical methods. We change several classical numerical methods to corresponding pseudo numerical methods and find that the pseudo linear multi-step method is the best in most situations. According to our experiments, by directly using pre-trained models on Cifar10, CelebA and LSUN, PNDMs can generate higher quality synthetic images with only 50 steps compared with 1000-step DDIMs (20x speedup), significantly outperform DDIMs with 250 steps (by around 0.4 in FID) and have good generalization on different variance schedules. Our implementation is available at https://github.com/luping-liu/PNDM.

cs.CV↗

arXiv2022

A Mini-Block Fisher Method for Deep Neural Networks

Deep neural networks (DNNs) are currently predominantly trained using first-order methods. Some of these methods (e.g., Adam, AdaGrad, and RMSprop, and their variants) incorporate a small amount of curvature information by using a diagonal matrix to precondition the stochastic gradient. Recently, effective second-order methods, such as KFAC, K-BFGS, Shampoo, and TNT, have been developed for training DNNs, by preconditioning the stochastic gradient by layer-wise block-diagonal matrices. Here we propose a "mini-block Fisher (MBF)" preconditioned gradient method, that lies in between these two classes of methods. Specifically, our method uses a block-diagonal approximation to the empirical Fisher matrix, where for each layer in the DNN, whether it is convolutional or feed-forward and fully connected, the associated diagonal block is itself block-diagonal and is composed of a large number of mini-blocks of modest size. Our novel approach utilizes the parallelism of GPUs to efficiently perform computations on the large number of matrices in each layer. Consequently, MBF's per-iteration computational cost is only slightly higher than it is for first-order methods. The performance of our proposed method is compared to that of several baseline methods, on both autoencoder and CNN problems, to validate its effectiveness both in terms of time efficiency and generalization power. Finally, it is proved that an idealized version of MBF converges linearly.

cs.LG↗

arXiv2022

GenerSpeech: Towards Style Transfer for Generalizable Out-Of-Domain Text-to-Speech

Style transfer for out-of-domain (OOD) speech synthesis aims to generate speech samples with unseen style (e.g., speaker identity, emotion, and prosody) derived from an acoustic reference, while facing the following challenges: 1) The highly dynamic style features in expressive voice are difficult to model and transfer; and 2) the TTS models should be robust enough to handle diverse OOD conditions that differ from the source data. This paper proposes GenerSpeech, a text-to-speech model towards high-fidelity zero-shot style transfer of OOD custom voice. GenerSpeech decomposes the speech variation into the style-agnostic and style-specific parts by introducing two components: 1) a multi-level style adaptor to efficiently model a large range of style conditions, including global speaker and emotion characteristics, and the local (utterance, phoneme, and word-level) fine-grained prosodic representations; and 2) a generalizable content adaptor with Mix-Style Layer Normalization to eliminate style information in the linguistic content representation and thus improve model generalization. Our evaluations on zero-shot style transfer demonstrate that GenerSpeech surpasses the state-of-the-art models in terms of audio quality and style similarity. The extension studies to adaptive style transfer further show that GenerSpeech performs robustly in the few-shot data setting. Audio samples are available at https://GenerSpeech.github.io/

eess.AS↗

arXiv2022

The C/M Ratio of AGB Stars in the Local Group Galaxies

The number ratio of carbon-rich to oxygen-rich asymptotic giant branch (AGB) stars (the so-called C/M ratio) is closely related to the evolution environment of the host galaxy. This work studies the C/M ratio in 14 galaxies within the Local Group with the most complete and clean sample of member stars identified in our previous works. The borderlines between carbon-rich AGB and oxygen-rich AGB stars as well as red supergiants are defined by Gaussian mixture model fitting to the number density in the $(J - K)/K$ diagram for the member stars of the LMC and M33, and then applied to the other galaxies by shifting the difference in the position of tip red giant branch (TRGB). The C/M ratios are obtained after precise and consistent categorization. Although for galaxies with larger distance modulo there is greater uncertainty, the C/M ratio is clearly found to decrease with the color index $(J - K)_0$ of TRGB as the indicator of metallicity, which agrees with previous studies and can be explained by the fact that carbon stars are more easily formed in a metal-poor environment. Furthermore, the C/M ratio within M33 is found to increase with galactocentric distance, which coincides with this scenario and the galactic chemical evolution model. On the other hand, the C/M ratio within M31 is found to decrease with galactocentric radius, which deserves further study.

astro-ph.GA↗

arXiv2022

Video-Guided Curriculum Learning for Spoken Video Grounding

In this paper, we introduce a new task, spoken video grounding (SVG), which aims to localize the desired video fragments from spoken language descriptions. Compared with using text, employing audio requires the model to directly exploit the useful phonemes and syllables related to the video from raw speech. Moreover, we randomly add environmental noises to this speech audio, further increasing the difficulty of this task and better simulating real applications. To rectify the discriminative phonemes and extract video-related information from noisy audio, we develop a novel video-guided curriculum learning (VGCL) during the audio pre-training process, which can make use of the vital visual perceptions to help understand the spoken language and suppress the external noise. Considering during inference the model can not obtain ground truth video segments, we design a curriculum strategy that gradually shifts the input video from the ground truth to the entire video content during pre-training. Finally, the model can learn how to extract critical visual information from the entire video clip to help understand the spoken language. In addition, we collect the first large-scale spoken video grounding dataset based on ActivityNet, which is named as ActivityNet Speech dataset. Extensive experiments demonstrate our proposed video-guided curriculum learning can facilitate the pre-training process to obtain a mutual audio encoder, significantly promoting the performance of spoken video grounding tasks. Moreover, we prove that in the case of noisy sound, our model outperforms the method that grounding video with ASR transcripts, further demonstrating the effectiveness of our curriculum strategy.

cs.CV↗

arXiv2022

FastSpeech 2: Fast and High-Quality End-to-End Text to Speech

Non-autoregressive text to speech (TTS) models such as FastSpeech can synthesize speech significantly faster than previous autoregressive models with comparable quality. The training of FastSpeech model relies on an autoregressive teacher model for duration prediction (to provide more information as input) and knowledge distillation (to simplify the data distribution in output), which can ease the one-to-many mapping problem (i.e., multiple speech variations correspond to the same text) in TTS. However, FastSpeech has several disadvantages: 1) the teacher-student distillation pipeline is complicated and time-consuming, 2) the duration extracted from the teacher model is not accurate enough, and the target mel-spectrograms distilled from teacher model suffer from information loss due to data simplification, both of which limit the voice quality. In this paper, we propose FastSpeech 2, which addresses the issues in FastSpeech and better solves the one-to-many mapping problem in TTS by 1) directly training the model with ground-truth target instead of the simplified output from teacher, and 2) introducing more variation information of speech (e.g., pitch, energy and more accurate duration) as conditional inputs. Specifically, we extract duration, pitch and energy from speech waveform and directly take them as conditional inputs in training and use predicted values in inference. We further design FastSpeech 2s, which is the first attempt to directly generate speech waveform from text in parallel, enjoying the benefit of fully end-to-end inference. Experimental results show that 1) FastSpeech 2 achieves a 3x training speed-up over FastSpeech, and FastSpeech 2s enjoys even faster inference speed; 2) FastSpeech 2 and 2s outperform FastSpeech in voice quality, and FastSpeech 2 can even surpass autoregressive models. Audio samples are available at https://speechresearch.github.io/fastspeech2/.

eess.AS↗