Popular stories.
Scientific research in the past 14 days

High intelligence: A risk factor for psychological and physiological overexcitabilities - ScienceDirect
www.sciencedirect.com

1 min read, listed 6 days ago in reddit/transhumanism/Gene therapy for human enhancement soon?

High intelligence is touted as being predictive of positive outcomes including educational success and income level. However, little is known about th…

[1610.00468] On the Modeling of Musical Solos as Complex Networks
arxiv.org

1 min read, listed 3 days ago in Giuseppe Sollazzo — 387: quantum of sollazzo

Notes in a musical piece are building blocks employed in non-random ways to create melodies. It is the "interaction" among a limited amount of notes that allows constructing the variety of musical compositions that have been written in centuries and within different cultures. Networks are a modeling tool that is commonly employed to represent a set of entities interacting in some way. Thus, notes composing a melody can be seen as nodes of a network that are connected whenever these are played in sequence. The outcome of such a process results in a directed graph. By using complex network theory, some main metrics of musical graphs can be meas...

The Venusian Lower Atmosphere Haze as a Depot for Desiccated Microbial Life: A Proposed Life Cycle for Persistence of the Venusian Aerial Biosphere | Astrobiology
www.liebertpub.com

52 min read, listed 3 days ago in MIT Technology Review — Gas spotted in Venus’s clouds could be a sign of alien life

Frontiers | The Easy Part of the Hard Problem: A Resonance Theory of Consciousness | Human Neuroscience
www.frontiersin.org

58 min read, listed 13 days ago in reddit/ArtificialInteligence/At which point do consciousness appear in a simulation?

Synchronization, harmonization, vibrations, or simply resonance in its most general sense seems to have an integral relationship with consciousness itself. One of the possible “neural correlates of consciousness” in mammalian brains is a specific combination of gamma, beta and theta electrical synchrony. More broadly, we see similar kinds of resonance patterns in living and non-living structures of many types. What clues can resonance provide about the nature of consciousness more generally? This paper provides an overview of resonating structures in the fields of neuroscience, biology and physics and offers a possible solution to what we see...

A comprehensive survey on model compression and acceleration | SpringerLink
link.springer.com

21 min read, listed 6 days ago in Latest Results for Artificial Intelligence Review

In recent years, machine learning (ML) and deep learning (DL) have shown remarkable improvement in computer vision, natural language processing, stock prediction, forecasting, and audio processing to name a few. The size of the trained DL model is large for these complex tasks, which makes it difficult to deploy on resource-constrained devices. For instance, size of the pre-trained VGG16 model trained on the ImageNet dataset is more than 500 MB. Resource-constrained devices such as mobile phones and internet of things devices have limited memory and less computation power. For real-time applications, the trained models should be deployed...

A comprehensive survey on model compression and acceleration | SpringerLink
link.springer.com

21 min read, listed 3 days ago in Latest Results for Artificial Intelligence Review

In recent years, machine learning (ML) and deep learning (DL) have shown remarkable improvement in computer vision, natural language processing, stock prediction, forecasting, and audio processing to name a few. The size of the trained DL model is large for these complex tasks, which makes it difficult to deploy on resource-constrained devices. For instance, size of the pre-trained VGG16 model trained on the ImageNet dataset is more than 500 MB. Resource-constrained devices such as mobile phones and internet of things devices have limited memory and less computation power. For real-time applications, the trained models should be deployed...

[2009.03393] Generative Language Modeling for Automated Theorem Proving
arxiv.org

1 min read, listed 9 days ago in reddit/MachineLearning/[R] GPT-f: a new SOTA for automated mathematical proofs

We explore the application of transformer-based language models to automated theorem proving. This work is motivated by the possibility that a major limitation of automated theorem provers compared to humans -- the generation of original mathematical terms -- might be addressable via generation from language models. We present an automated prover and proof assistant, GPT-f, for the Metamath formalization language, and analyze its performance. GPT-f found new short proofs that were accepted into the main Metamath library, which is to our knowledge, the first time a deep-learning based system has contributed proofs that were adopted by a formal...

[2008.11790] MutaGAN: A Seq2seq GAN Framework to Predict Mutations of Evolving Protein Populations
arxiv.org

2 min read, listed 16 hours ago in reddit/MachineLearning/[R] MutaGAN: A Seq2seq GAN Framework to Predict Mutations of Evolving Protein Populations

The ability to predict the evolution of a pathogen would significantly improve the ability to control, prevent, and treat disease. Despite significant progress in other problem spaces, deep learning has yet to contribute to the issue of predicting mutations of evolving populations. To address this gap, we developed a novel machine learning framework using generative adversarial networks (GANs) with recurrent neural networks (RNNs) to accurately predict genetic mutations and evolution of future biological populations. Using a generalized time-reversible phylogenetic model of protein evolution with bootstrapped maximum likelihood tree estimatio...

A systematic review of gamification techniques applied to elderly care | SpringerLink
link.springer.com

21 min read, listed 6 days ago in Latest Results for Artificial Intelligence Review

The proportion of the world’s population growing older is rapidly increasing over the last decades. With the recent progresses seen in information and communication technologies, there have been great concerns about developing personalized healthcare services that can ensure the living conditions and active aging of the elderly people. Among these technologies, we highlight and review in this work, the current state of gamification and related techniques applied to the elderly care context. Six research questions were defined to provide an overview on the current state in the development of gamified systems for elderly care through the identi...

A systematic review of gamification techniques applied to elderly care | SpringerLink
link.springer.com

21 min read, listed 3 days ago in Latest Results for Artificial Intelligence Review

The proportion of the world’s population growing older is rapidly increasing over the last decades. With the recent progresses seen in information and communication technologies, there have been great concerns about developing personalized healthcare services that can ensure the living conditions and active aging of the elderly people. Among these technologies, we highlight and review in this work, the current state of gamification and related techniques applied to the elderly care context. Six research questions were defined to provide an overview on the current state in the development of gamified systems for elderly care through the identi...

[2009.04374] Assessing Game Balance with AlphaZero: Exploring Alternative Rule Sets in Chess
arxiv.org

2 min read, listed 8 days ago in reddit/MachineLearning/[R] Assessing Game Balance with AlphaZero: Exploring Alternative Rule Sets in Chess

It is non-trivial to design engaging and balanced sets of game rules. Modern chess has evolved over centuries, but without a similar recourse to history, the consequences of rule changes to game dynamics are difficult to predict. AlphaZero provides an alternative in silico means of game balance assessment. It is a system that can learn near-optimal strategies for any rule set from scratch, without any human supervision, by continually learning from its own experience. In this study we use AlphaZero to creatively explore and design new chess variants. There is growing interest in chess variants like Fischer Random Chess, because of classical c...

Multiple Object Tracking | Papers With Code
paperswithcode.com

1 min read, listed 12 days ago in reddit/deeplearning/Please share a recent Multi Object Tracking code for Person tracking that they had good experience with

**Multiple Object Tracking** is the problem of automatically identifying multiple objects in a video and representing them as a set of trajectories with high accuracy. Source: [SOT for MOT ](https://arxiv.org/abs/1712.01059)

Simple linear regression - Wikipedia
en.wikipedia.org

17 min read, listed 5 days ago in reddit/artificial/Stupid question about some AI basics

[2009.07118] It's Not Just Size That Matters: Small Language Models Are Also Few-Shot Learners
arxiv.org

1 min read, listed 2 days ago in reddit/MachineLearning/[R] It's Not Just Size That Matters: Small Language Models Are Also Few-Shot Learners

When scaled to hundreds of billions of parameters, pretrained language models such as GPT-3 (Brown et al., 2020) achieve remarkable few-shot performance on challenging natural language understanding benchmarks. In this work, we show that performance similar to GPT-3 can be obtained with language models whose parameter count is several orders of magnitude smaller. This is achieved by converting textual inputs into cloze questions that contain some form of task description, combined with gradient-based optimization; additionally exploiting unlabeled data gives further improvements. Based on our findings, we identify several key factors required...

[1609.00408] Defeating Image Obfuscation with Deep Learning
arxiv.org

1 min read, listed 9 days ago in reddit/MachineLearning/[PROJECT] Cover Faces from Protest Photos

We demonstrate that modern image recognition methods based on artificial neural networks can recover hidden information from images protected by various forms of obfuscation. The obfuscation techniques considered in this paper are mosaicing (also known as pixelation), blurring (as used by YouTube), and P3, a recently proposed system for privacy-preserving photo sharing that encrypts the significant JPEG coefficients to make images unrecognizable by humans. We empirically show how to train artificial neural networks to successfully identify faces and recognize objects and handwritten digits even if the images are protected using any of the abo...

Google AI Blog: Improving the Accuracy of Genomic Analysis with DeepVariant 1.0
ai.googleblog.com

8 min read, listed 10 hours ago in Google AI Blog

Posted by Andrew Carroll, Product Lead, and Pi-Chuan Chang, Technical Lead, Google Health Sequencing genomes involves sampling short piec...

[2009.05451] A Comparison of LSTM and BERT for Small Corpus
arxiv.org

1 min read, listed 3 days ago in reddit/MachineLearning/[R] A Comparison of LSTM and BERT for Small Corpus: “Our experimental results show that bidirectional LSTM models can achieve significantly higher results than a BERT model for a small dataset and these simple models get trained in much less time than tuning the pretrained counterparts.”

Recent advancements in the NLP field showed that transfer learning helps with achieving state-of-the-art results for new tasks by tuning pre-trained models instead of starting from scratch. Transformers have made a significant improvement in creating new state-of-the-art results for many NLP tasks including but not limited to text classification, text generation, and sequence labeling. Most of these success stories were based on large datasets. In this paper we focus on a real-life scenario that scientists in academia and industry face frequently: given a small dataset, can we use a large pre-trained model like BERT and get better results tha...

Google AI Blog: Improving Sparse Training with RigL
ai.googleblog.com

6 min read, listed 2 days ago in Google AI Blog

Posted by Utku Evci and Pablo Samuel Castro, Research Engineers, Google Research, Montreal Modern deep neural network architectures are o...

[2009.06489] The Hardware Lottery
arxiv.org

1 min read, listed 1 day ago in reddit/MachineLearning/[R] The Hardware Lottery: “The advent of domain specialized hardware makes it increasingly costly to stray off of the beaten path of research ideas.” (Essay by Sara Hooker)

Hardware, systems and algorithms research communities have historically had different incentive structures and fluctuating motivation to engage with each other explicitly. This historical treatment is odd given that hardware and software have frequently determined which research ideas succeed (and fail). This essay introduces the term hardware lottery to describe when a research idea wins because it is suited to the available software and hardware and not because the idea is superior to alternative research directions. Examples from early computer science history illustrate how hardware lotteries can delay research progress by casting success...

[2002.10698] Hierarchical Conditional Relation Networks for Video Question Answering
arxiv.org

1 min read, listed 5 days ago in reddit/MachineLearning/[R] Hierarchical Conditional Relation Networks (HCRN) for Video Question Answering - CVPR 2020 (link to free Zoom lecture by the authors in comments)

Video question answering (VideoQA) is challenging as it requires modeling capacity to distill dynamic visual artifacts and distant relations and to associate them with linguistic concepts. We introduce a general-purpose reusable neural unit called Conditional Relation Network (CRN) that serves as a building block to construct more sophisticated structures for representation and reasoning over video. CRN takes as input an array of tensorial objects and a conditioning feature, and computes an array of encoded output objects. Model building becomes a simple exercise of replication, rearrangement and stacking of these reusable units for diverse m...

AWAC: Accelerating Online Reinforcement Learning with Offline Datasets – The Berkeley Artificial Intelligence Research Blog
bair.berkeley.edu

12 min read, listed 9 days ago in Berkeley AI Research

The BAIR Blog

[2009.06732] Efficient Transformers: A Survey
arxiv.org

1 min read, listed 2 days ago in reddit/MachineLearning/[R] Efficient Transformers: A Survey

Transformer model architectures have garnered immense interest lately due to their effectiveness across a range of domains like language, vision and reinforcement learning. In the field of natural language processing for example, Transformers have become an indispensable staple in the modern deep learning stack. Recently, a dizzying number of "X-former" models have been proposed - Reformer, Linformer, Performer, Longformer, to name a few - which improve upon the original Transformer architecture, many of which make improvements around computational and memory efficiency. With the aim of helping the avid researcher navigate this flurry, this p...

Illuminating the dark spaces of healthcare with ambient intelligence | Nature
www.nature.com

8 min read, listed 9 days ago in Twitter — recent in #ai

Breakthroughs in artificial intelligence and low-cost, contactless sensors have given rise to an ambient intelligence that can potentially improve the physical execution of healthcare delivery, if used in a thoughtful manner.

Bottom-up fabrication of nanoelectronics with DNA nanotechnology
www.nanowerk.com

4 min read, listed 5 days ago in reddit/singularity/Bottom-up fabrication of nanoelectronics with DNA nanotechnology

DNA is probably the most programmable biomaterial for creating a wide range of rationally designed and functionally enhanced nanostructures. The sophisticated, programmable, and addressable DNA nanostructures are strong candidates for constructing nanoelectronic devices. The size of DNA molecules is also key: DNA double-helix has a neighboring base pair distance of 0.34 nm and a diameter of 2.1 - 2.6 nm a, and thus DNA complex-based nanoelectronics may break the 5-nm processing limit of commercial silicon-based semiconductors.

Human3.6M Dataset
vision.imar.ro

2 min read, listed 6 days ago in reddit/datasets/Is Human3.6m dead?