NP-Easy

Part time computer
Full time MPhil student at @ICRAR @UWAresearch
All time fan of humanity
Definitely one person and not several
Github: pritchardn

[NP-Easy @NicP42]

From the ArXiv: 26 https://t.co/LDjkecHwyM
A discussion of the graph isomorphism problem. One of the most peculiar classic computer science problems.
Thanks Martin Grohe and Daniel Neuen
#FTArXiV #graphs

[NP-Easy @NicP42]

From the ArXiv: 25 https://t.co/SUI3hGvctM
Fast and accurate prediction of pedestrian behaviour is essential for driverless cars to succeed.
Thanks Prateek Agrawal and @prprbr at @volkswagen
#FTArXiV #AI #driverless

[NP-Easy @NicP42]

From the ArXiv: 24 https://t.co/NTZoxqPZPg
A concise introduction to levels of artificial intelligence and human-like language acquisition. Perhaps one day we might raise AI like we do children.
Thanks @intuinno
#FTArXiV #AI

[NP-Easy @NicP42]

From the ArXiv: 23 https://t.co/thZrBBaOMH
An automated dog-trainer powered by a #NVIDIA Jetson Nano. I wonder if it could work on people? (asking for a friend)
Thanks Jason Stock and Tom Cavey @ResearchCSU
#FTArXiV #dogs #reinforcementlearning

[NP-Easy @NicP42]

From the ArXiv: 22 https://t.co/l2g3u349AV
An impressive approach to correcting projections onto distorted surfaces.
Thanks Bingyao Huang, Tao Sun and Haibin Ling
#FTArXiV #projection

[NP-Easy @NicP42]

From the ArXiv: 21 https://t.co/K8LZgtq9hL
Novel simulations better model globular clusters with good comparison to physical observations.
Thanks @MadellanicCloud
#FTArXiV #47_TUC #globularclusters

[NP-Easy @NicP42]

From the ArXiv: 20 https://t.co/NfDnG4ZbRJ
IEEE floating point arithmetic is not perfect, posits are less error prone and lead to greater machine learning performance.
Thanks @ihstein et al.
#FTArXiV #computing #MachineLeraning

[NP-Easy @NicP42]

From the ArXiv: 19 https://t.co/b7KFHWoFYM
High-level quantum computing programming needs scalable error optimising compilers. Here is a prototype approach using near-symbolic expressions
Thanks Guilia Meuli @msoeken @MartinQuantum and Thomas Häner
#FTArXiV #QuantumComputing

[NP-Easy @NicP42]

From the ArXiv: 18 https://t.co/3n8y3rm9uP
A practical review of deep learning in electron microscopy. For those interested here's a good place to start learning.
Thanks Jeffrey M. Ede from @warwickuni
#FTArXiV #DeepLearning #electronmicroscopy

[NP-Easy @NicP42]

From the ArXiv: 17 https://t.co/ARln5OaPDX
An optimization on quantum circuit simulation by adding a specialized compiler pass, typically halving simulating time on random circuits.
Thanks @GianGuerreschi from @intel
#FTArXiV #QuantumComputing

[NP-Easy @NicP42]

From the ArXiv: 16 https://t.co/cEVzkzB9mh
A new distributed consensus protocol based on coding theory and not cryptography for its security, achieving correctness and optimal characteristics against unbounded adversaries.
Thanks Jinyuan Chen from @LATech
#FTArXiV #Byzantine #ECC

[NP-Easy @NicP42]

From the ArXiv: 15 https://t.co/fhi6MWKMxk
The amount of information a word carries correlates with its phonetic characteristics in Mandarin Chinese, a trend shared with the English language.
Thanks @tang_kevin and Jason Shaw

#FTArXiV #prosody #linguistics

[NP-Easy @NicP42]

From the ArXiv: 14 https://t.co/WhAx41DOAp
An entertaining introduction to the arguments between computation and causal cognition.
Perhaps truly comprehensive AI will forever be just out of reach.
Thanks @ProfJMarkBishop
#FTArXiV #ArtificialIntelligence #NeuralNetworks

[NP-Easy @NicP42]

From the ArXiv: 13 https://t.co/2Anb6L1sb1
Dialogue contradictions are obvious flaws when dealing with chat-bot-style software. Here's a new dataset and approach to improving the state-of-the-art in chatbots.
Thanks @EasonNie et al. @UNC @facebookai
#FTArXiV #chatbot

[NP-Easy @NicP42]

From the ArXiv: 12 https://t.co/Rgp35LIVqj
A simple, intuitive and performant automatic caption generation metric, a neat baseline to judge more complex methods against.
Thanks Naeha Sharif Lyndon White @MBennamoun Wei Liu and @SyedAfaqAShah
#FTArXiV

[NP-Easy @NicP42]

From the ArXiv: 11 https://t.co/b9xv0xS9zc
Superior spam spotting with GPT-2 when labelled data is scarce.
Cheers Athirai A. Irissappane et al. @uwtacoma
#GPT2 #Spam #DeepLearning
#FTArXiV

[NP-Easy @NicP42]

From the ArXiv: 10 https://t.co/cPcAphpa59
A Chrismissy take on the history of Medical AI Research.
All I want for Christmas is fully-transparent AI.
Thank you @tpq__ et al. @Deakin @unimelb
#MedTech #ethics #Christmas #FTArXiV

[NP-Easy @NicP42]

From the ArXiv 9: https://t.co/6mZjc1JjZt
Here is a distributed connected component finding algorithm working at scales beyond any other approach.
100 Billion nodes is no joke! And from Walmart! What are you all up to?
Thanks, Saigopal Thota et al. @Walmartlabs
#Graphs #FTArXiv

[NP-Easy @NicP42]

From the ArXiv 8: https://t.co/qv8k96leb3
Using neural networks to identify foods for nutrition-tracking.
Even if it gets close to the real answer, this will save so much time!
Thank you Charles Freitas et al. @UFRPEOnline
#NeuralNetworks #FTArXiV

[NP-Easy @NicP42]

From the ArXiv 7: https://t.co/DUKKXtFC0U
A lightweight physically-based renderer. Better renders in the same time as Phong based methods with a simple Python interface.
Thank you @alejandro_rosu and @SvenBehnke
#rendering #FTArXiv