49. AI Used to Detect Deepfakes, NASA’s Airless Tires Hitting The Market, VR Utilized in Retinal Prosthetics

49. AI Used to Detect Deepfakes, NASA’s Airless Tires Hitting The Market, VR Utilized in Retinal Prosthetics
Cool STEM News:
Scientists developed a clever way to detect Deepfakes by analyzing light reflections in the eyes | The Next Web (01:45)
- Deepfakes are always mentioned as a future technology that people are worried about.
- Deepfakes: artificial images and sounds put together with machine-learning algorithms to manipulate how a video looks and sounds.
- Main worry is that it can make it look like real people said/did something they didn’t.
- But what is not mentioned all that often, because we focus on the negative more, is the effort to counteract these “bad guy” technologies.
- Computer Scientists at the University of Buffalo developed a method to detect these Deepfake videos with a 94% accuracy.
- Note: Accuracy based on portrait style photos
- How? By utilizing light reflected in the eyes.
- Analyzed the corneas, which generate reflective patterns when illuminated by light.
- Exposed to the same light type your corneas will reflect the same pattern.
- Deepfakes since they are GAN-synthesized (machine-learning algorithm) the images will fail to capture the light reflection properly and show inconsistencies.
- GAN (Generative adversarial network): In my words, is a machine-learning algorithm that is trained on whether or not it fools the observer.
- The researchers plan on working out the shortcomings found throughout the study:
- Needs two eyes refracting light
- One eye not visible it won’t work.
- Only proven effective on portrait style images
New Clues Uncovered about Toxic Protein Structures in Neurodegenerative Diseases | GenEngNews (10:16)
- Transactive response DNA-binding protein-43 (TDP-43) is a soluble protein that interacts with nucleic acids.
- The issue is that this protein forms large clumps, known as amyloid fibrils, which are associated with neurological diseases.
- ALS, Alzheimer’s, sclerosis, dementia, and Huntington’s disease
- Scientists at the Case Western Reserve University School of Medicine, located in Ohio, were able to determine the structures of TDP-43.
- Analyzed thousands of images of fibrils formed in the test tube by the key fragment of TDP-43.
- Resolution close to individual atoms
- This understanding provides insight into how these toxic proteins clump and spread between nerve cells in the brain.
- Qiuye Li, a graduate student and lead author hits the above point: “This is really an exciting development because it reveals a mechanism for the growth of these toxic aggregates. This, in turn, provides important clues as to how these aggregates may spread between the cells in affected brains.”
- Could pave the way for developing new therapeutics to treat neurological diseases.
Genetic Scalpel Invented for Microalgae to Efficiently Turn CO2 Into Biofuel | SciTechDaily (16:21)
- Researchers from the Qingdao Institute of BioEnergy and Bioprocess Technology (QIBEBT) have genetically modified an oil producing microalgae to create a more efficient genome.
- In the process, they created a gene scalpel that can trim microalgal genomes “rapidly and creatively.”
- This modified microalgae is considered a “minimal genome”: a genome stripped of all duplicated or apparently non-functional genes (“junk genes”).
- Normally this minimizing process is done in research on more simplistic organisms not something like an algae (e.g. eukaryotic organisms).
- Eukaryote: any cell or organism that possesses a clearly defined nucleus.
- Researchers were able to accomplish targeted deletion of genes for a type of algae called Nannochloropsis oceanica.
- Single-Cell algae
- Potential for production of biofuels, biomaterials, and other chemicals.
- Cut out two low-expression regions (LER) — over 200 kilobases of the algae.
- Kilobase: length of a DNA equal to 1,000 base pairs
- Going Forward:
- Researchers want to see if they can snip out still further LERs and other non-lethal regions.
- Goal to craft a fully minimal microalgae that makes biofuels from CO2 with the highest efficiency.
- Other Benefits:
- Model organism for further study of the molecular and biological function of every gene
- Used as a building block for synthetic biologists for customized production of biomolecules such as biofuels or bioplastics
METL Tires Using NASA Tech Are Coming to Bikes | Gizmodo & New Atlas (24:21)
- Due to rovers exploring rough terrain on other planets, NASA had to develop a tire that was flexible enough & tough enough.
- Lead to the development of an airless titanium tire that’s flexible like rubber, but nearly indestructible.
- Nitinol: The metal alloy developed made of nickel and titanium.
- Exhibits the shape memory effect: when deformed it returns to its original manufactured shape without permanent damage.
- The first consumer-oriented application of airless shape memory alloy (SMA) tire technology will be developed by a startup called The Smart Tire Company.
- Announced that it’s creating a metal bicycle tire using NASA’s Nitinol alloy that never needs to be inflated.
- The creators are hopeful to get these tires out to market by 2022.
- “Cyclists will not be able to wait to get their hands on these very cool-looking, space-age Metl tires that don’t go flat,” says Earl Cole, CEO of The Smart Tire Company. “The unique combination of these advanced materials, coupled with a next generation, eco-friendly design make for a revolutionary product.”
- Spin-offs from NASA:
- Camera Phones: 1990s, JPL created a camera small enough to fit on a spacecraft (⅓ of all cameras contain this tech)
- Scratch Resistant Lens: Developed the technique utilize in the process
- Athletic Shoes
- LEDs
- CAT Scans
- Freeze Dried Foods
Quest for prosthetic retinas progresses toward human trials, with a VR assist | TechCrunch (32:31)
- Latest advancement was accomplished by Diego Ghezzi, who holds the Medtronic Chair in Neuroengineering (LNE) at EPFL’s School of Engineering.
- Located in Lausanne, Switzerland
- “Our system is designed to give blind people a form of artificial vision by using electrodes to stimulate their retinal cells,” says Ghezzi.
- Process from EPFL News:
- Captures field of view image from an outside device (e.g. glasses)
- Sends the data to a microcomputer placed in one of the eyeglasses.
- Microcomputer turns the data into light signals which are transmitted to electrodes in the retinal implant.
- Electrodes then stimulate the retina in such a way that the wearer sees a simplified, black-and-white version of the image.
- Wearers must learn to interpret the many dots of light (10,500) in order to make out shapes and objects.
- Ghezzi gives an analogy as to what the wearer would see: “It’s like when you look at stars in the night sky – you can learn to recognize specific constellations. Blind patients would see something similar with our system.”
- To test this somehow the researchers couldn’t just install a retinal implant on a human participant, so they went to a medium you may not have expected: VR.
- Simulated in a VR environment what a person with this retinal implant would see:
- Dark except for little simulated “phosphors,” the pinpricks of light they expect to create by stimulating the retina.
- Primary finding: Most important factor was visual angle — the overall size of the area where the image appears.
- This demonstration showed that the implant’s parameters are theoretically sound and the team can start working toward human trials.
Support The Podcast:
Website: https://thatscoolnews.com/
Support the Podcast: https://thatscoolnews.com/support
Review The Podcast: https://thatscoolnews.com/review/
Email: [email protected]
Follow On Social Media:
Instagram: https://www.instagram.com/thatscoolnews/
Telegram: https://t.me/thatscoolnews