Papers Read on AI

While there have been recent advances in few-shot image stylization, these methods fail to capture stylistic details that are obvious to humans. Details such as the shape of the eyes, the boldness of the lines, are especially difficult for a model to learn, especially so under a limited data setting. In this work, we aim to perform oneshot image stylization that gets the details right.

2021: Min Jin Chong, David Forsyth

https://arxiv.org/pdf/2112.11641v1.pdf

We show that diffusion models can achieve image sample quality superior to the current state-of-the-art generative models. We achieve this on unconditional image synthesis by finding a better architecture through a series of ablations.

2021: Prafulla Dhariwal, Alex Nichol

https://arxiv.org/pdf/2105.05233v4.pdf

Recent work has shown that self-supervised pre-training leads to improvements over supervised learning on challenging visual recognition tasks. CLIP, an exciting new approach to learning with language supervision, demonstrates promising performance on a wide variety of benchmarks. In this work, we explore whether self-supervised learning can aid in the use of language supervision for visual representation learning. We introduce SLIP, a multi-task learning framework for combining self-supervised learning and CLIP pre-training.

2021: Norman Mu, Alexander Kirillov, David A. Wagner, Saining Xie

https://arxiv.org/pdf/2112.12750v1.pdf

Recent work has shown that self-supervised pre-training leads to improvements over supervised learning on challenging visual recognition tasks. CLIP, an exciting new approach to learning with language supervision, demonstrates promising performance on a wide variety of benchmarks. In this work, we explore whether self-supervised learning can aid in the use of language supervision for visual representation learning. We introduce SLIP, a multi-task learning framework for combining self-supervised learning and CLIP pre-training.

2021: Norman Mu, Alexander Kirillov, David A. Wagner, Saining Xie

https://arxiv.org/pdf/2112.12750v1.pdf

Recent work has shown that self-supervised pre-training leads to improvements over supervised learning on challenging visual recognition tasks. CLIP, an exciting new approach to learning with language supervision, demonstrates promising performance on a wide variety of benchmarks. In this work, we explore whether self-supervised learning can aid in the use of language supervision for visual representation learning. We introduce SLIP, a multi-task learning framework for combining self-supervised learning and CLIP pre-training.

2021: Norman Mu, Alexander Kirillov, David A. Wagner, Saining Xie

https://arxiv.org/pdf/2112.12750v1.pdf

Data augmentation is an important component in the robustness evaluation of models in natural language processing (NLP) and in enhancing the diversity of the data they are trained on. In this paper, we present NL-Augmenter, a new participatory Pythonbased natural language augmentation framework which supports the creation of both transformations (modifications to the data) and filters (data splits according to specific features). We describe the framework and an initial set of 117 transformations and 23 filters for a variety of natural language tasks.

2021: Kaustubh D. Dhole, Varun Gangal, Sebastian Gehrmann, Aadesh Gupta, Zhenhao Li, Saad Mahamood, Abinaya Mahendiran, Simon Mille, Ashish Srivastava, Samson Tan, Tongshuang (Sherry) Wu, J. Sohl-Dickstein, Jinho D. Choi, E. Hovy, Ondrej Dusek, Sebastian Ruder, Sajant Anand, Nagender Aneja, Rabin Banjade, Lisa Barthe, Hanna Behnke, Ian Berlot-Attwell, Connor Boyle, C. Brun, Marco Antonio Sobrevilla Cabezudo, Samuel Cahyawijaya, E. Chapuis, Wanxiang Che, Mukund Choudhary, C. Clauss, Pierre Colombo, Filip Cornell, Gautier Dagan, Mayukh Das, Tanay Dixit, Thomas Dopierre, Paul-Alexis Dray, Suchitra Dubey, Tatiana Ekeinhor, Marco Di Giovanni, Rishabh Gupta, Louanes Hamla, Sang Han, Fabrice Harel-Canada, Antoine Honoré, Ishan Jindal, Przemyslaw K. Joniak, Denis Kleyko, Venelin Kovatchev, Kalpesh Krishna, Ashutosh Kumar, Stefan Langer, Seungjae Ryan Lee, Corey J. Levinson, Hualou Liang, Kaizhao Liang, Zhexiong Liu, Andrey Lukyanenko, V. Marivate, Gerard de Melo, Simon Meoni, Maxime Meyer, Afnan Mir, N. Moosavi, Niklas Muennighoff, Timothy Sum Hon Mun, Kenton W. Murray, M. Namysl, Maria Obedkova, Priti Oli, Nivranshu Pasricha, J. Pfister, R. Plant, Vinay Uday Prabhu, V. Pais, Libo Qin, Shahab Raji, Pawan Kumar Rajpoot, Vikas Raunak, Roy Rinberg, Nicolas M. Roberts, Juan Diego Rodriguez, C. Roux, S. VasconcellosP.H., Ananya B. Sai, Robin M. Schmidt, Thomas Scialom, T. Sefara, Saqib Shamsi, Xu-dong Shen, Haoyue Shi, Yiwen Shi, Anna V. Shvets, Nick Siegel, Damien Sileo, Jamie Simon, Chandan Singh, Roman Sitelew, P. Soni, Taylor M Sorensen, W. Soto, Aman Srivastava, K V Aditya Srivatsa, Tony Sun, T. MukundVarma, A. Tabassum, Fiona Anting Tan, Ryan Teehan, Monalisa Tiwari, Marie Tolkiehn, Athena Wang, Zijian Wang, Gloria Wang, Zijie Jay Wang, Fuxuan Wei, Bryan Wilie, Genta Indra Winata, Xinyi Wu, Witold Wydma'nski, Tianbao Xie, Usama Yaseen, M. Yee, Jing Zhang, Yue Zhang

https://arxiv.org/abs/2112.02721v1

This paper presents a grounded language-image pretraining (GLIP) model for learning object-level, languageaware, and semantic-rich visual representations.

2021: Liunian Harold Li, Pengchuan Zhang, Haotian Zhang, Jianwei Yang, Chunyuan Li, Yiwu Zhong, Lijuan Wang, Lu Yuan, Lei Zhang, Jenq-Neng Hwang, Kai-Wei Chang, Jianfeng Gao

Ranked #1 on Phrase Grounding on Flickr30k Entities Test (using extra training data)

https://arxiv.org/pdf/2112.03857v1.pdf

Neural vocoders, used for converting the spectral representations of an audio signal to the waveforms, are a commonly used component in speech synthesis pipelines. It focuses on synthesizing waveforms from low-dimensional representation, such as Mel-Spectrograms. In recent years, different approaches have been introduced to develop such vocoders. However, it becomes more challenging to assess these new vocoders and compare their performance to previous ones. To address this problem, we present VocBench, a framework that benchmark the performance of state-of-the art neural vocoders.

2021: Ehab A. AlBadawy, Andrew Gibiansky, Qing He, Jilong Wu, Ming-Ching Chang, Siwei Lyu

https://arxiv.org/pdf/2112.03099v1.pdf

This paper shows that masked autoencoders (MAE) are scalable self-supervised learners for computer vision. Our MAE approach is simple: we mask random patches of the input image and reconstruct the missing pixels. It is based on two core designs.

2021: Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Doll'ar, Ross B. Girshick

Ranked #1 on Image Classification on iNaturalist 2019

https://arxiv.org/pdf/2111.06377v2.pdf

YourTTS brings the power of a multilingual approach to the task of zero-shot multi-speaker TTS. Our method builds upon the VITS model and adds several novel modifications for zeroshot multi-speaker and multilingual training. We achieved stateof-the-art (SOTA) results in zero-shot multi-speaker TTS and results comparable to SOTA in zero-shot voice conversion on the VCTK dataset.

2021: Edresson Casanova, Julian Weber, C. Shulby, Arnaldo Cândido Júnior, Eren Gölge, M. Ponti

https://arxiv.org/pdf/2112.02418v1.pdf