AI-powered banana diseases and pest detection

Paper: https://plantmethods.biomedcentral.com/articles/10.1186/s13007-019-0475-z

These researchers have developed a tool to tackle these silent killers with an app that can scan banana plants for early signs of infection, and alert farmers before it takes hold on their crops.

A mathematical model from 103 years ago predicted something that was seen for the first time today: a #black_hole.

#MachineLearning could never do that: it needs observations to model anything. This is a major weak-point of ML. Let's fix it.

A stark contrast between Machine Learning vs other forms of mathematical modeling is that ML models often don't model extreme corner cases very well, because #data in those areas is rare. Gathering data in important areas is as important a skill as building fancy neural networks.

Sadly, too often, using extreme inputs to a model is more useful: e.g. by modeling physics of levers on light objects with short levers, we then built very long levers to lift extremely heavy things. Instead, ML is better suited at modeling everyday phenomena with complex models.

Blake Richards: Deep Learning with Ensembles of Neocortical Microcircuits


An inspiring talk by Blake Richards at the ICLR2018 emphasizing the interaction between neuroscience and machine learning. This intersection is where great things happen.

https://goo.gl/1YCjrm


https://t.iss.one/ArtificialIntelligenceArticles

NeurIPS 2019 Stats
Blog by Diego Charrez : https://medium.com/@dcharrezt/neurips-2019-stats-c91346d31c8f
#MachineLearning #Neurips #Neurips2019 #ArtficialIntelligence

Google today open-sourced Coached Conversational Preference Elicitation (CCPE) and Taskmaster-1, datasets of dialog between two people. Both datasets are being shared by Google AI researchers to supply the training material necessary to model natural language systems that achieve human-level performance.
https://venturebeat.com/2019/09/06/google-open-sources-datasets-for-ai-assistants-with-human-level-understanding/

Robotics in neuroscience

Robotics has many tools that can applied to neuro-prosthetics. For example, I research noninvasive neural interfaces for prosthetics control. For me focusing on learning perception (sensors), control techniques, and AI/machine was very helpful to develop control algorithms for physical devices. Signal processing is critical in my research as well.

The basics of kinematics / Dynamics is also helpful for studying human Biomechanics and analysing how the devices you design affect the users.

Check out the abstracts at this conference:

https://icorr2019.org/

It will give you an idea of what's happening at the intersection of robotics and rehab medicine.

The deepest problem with deep learning
Some reflections on an accidental Twitterstorm, the future of AI and deep learning, and what happens when you confuse a schoolbus with a snow plow.

https://medium.com/@GaryMarcus/the-deepest-problem-with-deep-learning-91c5991f5695

AutoGMM: Automatic Gaussian Mixture Modeling in Python. https://arxiv.org/abs/1909.02688

Intensity augmentation for domain transfer of whole breast segmentation in MRI. https://arxiv.org/abs/1909.02642

Deep Iterative Frame Interpolation for Full-frame Video Stabilization. https://arxiv.org/abs/1909.02641

Investigating BERT's Knowledge of Language: Five Analysis Methods with NPIs. https://arxiv.org/abs/1909.02597

TOP Paper

Detecting Deep Neural Network Defects with Data Flow Analysis. https://arxiv.org/abs/1909.02190

FastSpeech: Fast, Robust and Controllable Text to Speech
Ren et al.: https://arxiv.org/abs/1905.09263
Samples : https://speechresearch.github.io/fastspeech/
#ArtificialIntelligence #DeepLearning #MachineLearning

In eras of technological disruption, leadership matters.


Andrew Ng speaks about the progress of AI, how to accelerate AI adoption, and what’s around the corner for AI at Amazon re:MARS 2019 in Las Vegas, California.

https://www.youtube.com/watch?v=j2nGxw8sKYU

On Extractive and Abstractive Neural Document Summarization with Transformer Language Models"
Sandeep Subramanian, Raymond Li, Jonathan Pilault, Christopher Pal : https://arxiv.org/abs/1909.03186
#transformer #naturallanguageprocessing #machinelearning

People should stop training radiologists now. It is just completely obvious within 5 years deep learning is going to do better than radiologists."

-- Geoffrey Hinton

https://www.youtube.com/watch?v=2HMPRXstSvQ&feature=youtu.be&t=29