✨💡 This is an ace idea from @sarah_edo! 💕
👩💻 Be on the lookout for a @TensorFlow Advent Calendar tomorrow, as well, highlighting meaningful, high-impact projects and papers from our community. If you'd like for yours to be considered, please shoot me an @-mention!
👩💻 Be on the lookout for a @TensorFlow Advent Calendar tomorrow, as well, highlighting meaningful, high-impact projects and papers from our community. If you'd like for yours to be considered, please shoot me an @-mention!
✨#TFadvent begins today! 😄
For our first project, I'd like to highlight this accessibility example from @shekitup that uses @TensorFlowJS to interpret sign language—and then translates those signs into input that can be used by a home assistant! 🗣️✨
medium.com/tensorflow/get…
For our first project, I'd like to highlight this accessibility example from @shekitup that uses @TensorFlowJS to interpret sign language—and then translates those signs into input that can be used by a home assistant! 🗣️✨
medium.com/tensorflow/get…
🎁 Day #2 of #TFadvent:
🤖 Check out this project from a recent hackathon at @ColbyCollege! The team trained a @TensorFlow model to learn muscle movements, then used that model to send signals to a prosthetic arm, controlling one finger at a time. ✨
👉: m.facebook.com/notes/daviscon…
🤖 Check out this project from a recent hackathon at @ColbyCollege! The team trained a @TensorFlow model to learn muscle movements, then used that model to send signals to a prosthetic arm, controlling one finger at a time. ✨
👉: m.facebook.com/notes/daviscon…
🎉 Day #3 of #TFadvent:
#Creatability, a collection of experiments exploring ways to make creative tools more accessible. The team: @AbilityNYU @JAZDeafMusical @openupmusic @ChanceyFleet @BerkeleyBlink, and more.
😁 Many are built with @TensorFlowJS!
👉 g.co/creatability
#Creatability, a collection of experiments exploring ways to make creative tools more accessible. The team: @AbilityNYU @JAZDeafMusical @openupmusic @ChanceyFleet @BerkeleyBlink, and more.
😁 Many are built with @TensorFlowJS!
👉 g.co/creatability
🧙♂️ Day #4 of #TFadvent:
A #VR app from @jamesckane made with @unity3d, @LeapMotion hand-tracking, and @TensorFlow! Predicts 3D glyph drawings and casts specific magic spells in response. ✨
👉 Learn more about the project and how to make your own magic: jamesckane.com/case-study/app…
A #VR app from @jamesckane made with @unity3d, @LeapMotion hand-tracking, and @TensorFlow! Predicts 3D glyph drawings and casts specific magic spells in response. ✨
👉 Learn more about the project and how to make your own magic: jamesckane.com/case-study/app…
✨🎹 Day #5 of @TFadvent:
Check out Latent Cycles from @teropa! Each of the four corners in this space is a little melodic pattern generated using ImprovRNN and #Magenta.
🖊️ Run this example in a browser with @CodePen: codepen.io/teropa/full/rd…
📰 Blog: magenta.tensorflow.org/blog/2018/05/0…
Check out Latent Cycles from @teropa! Each of the four corners in this space is a little melodic pattern generated using ImprovRNN and #Magenta.
🖊️ Run this example in a browser with @CodePen: codepen.io/teropa/full/rd…
📰 Blog: magenta.tensorflow.org/blog/2018/05/0…
✨🥔 Day #6 of #TFadvent:
Inspecting and sorting good ingredients from bad is one of the most important tasks in food processing—but that task can be time consuming and prone to error.
👉 Learn how @TensorFlow+@KewpieUSA are partnering for food quality:
blog.google/products/googl…
Inspecting and sorting good ingredients from bad is one of the most important tasks in food processing—but that task can be time consuming and prone to error.
👉 Learn how @TensorFlow+@KewpieUSA are partnering for food quality:
blog.google/products/googl…
#TFadvent Day #7
From the "Machine Learning for Creativity and Design" Workshop at #NeurIPS2018 tomorrow:
🖼️ SpaceSheets: a general-purpose @TensorFlowJS spreadsheet interface designed to support the experimentation and exploration of latent spaces.
👉 vusd.github.io/spacesheet/
From the "Machine Learning for Creativity and Design" Workshop at #NeurIPS2018 tomorrow:
🖼️ SpaceSheets: a general-purpose @TensorFlowJS spreadsheet interface designed to support the experimentation and exploration of latent spaces.
👉 vusd.github.io/spacesheet/
#TFadvent Day #8
👩🔬 CASP is a virtual protein-folding Olympics, where the aim is to predict 3D structures of proteins based on genetic sequence data.
This has huge implications for health-related research (Alzheimer's, Parkinson's, Huntington's, etc.).
theguardian.com/science/2018/d…
👩🔬 CASP is a virtual protein-folding Olympics, where the aim is to predict 3D structures of proteins based on genetic sequence data.
This has huge implications for health-related research (Alzheimer's, Parkinson's, Huntington's, etc.).
theguardian.com/science/2018/d…
✨🧪 Last week, @DeepMindAI's AlphaFold SUMZ score of 127.99 was 20 points higher than the 2nd-ranked team - achieving what CASP called “unprecedented progress in the ability of computational methods to predict protein structure.”
👉 Read more here: medium.com/syncedreview/d…
👉 Read more here: medium.com/syncedreview/d…
✨👩🔬 "We are especially excited about how it might improve our understanding of the body and how it works, enabling scientists to design new, effective cures for diseases [like Parkinson's] more efficiently."
❤️.
🧠 @DeepMindAI's #AlphaFold write-up: deepmind.com/blog/alphafold/
❤️.
🧠 @DeepMindAI's #AlphaFold write-up: deepmind.com/blog/alphafold/
#TFadvent Day #9
👫 Autism impacts >1 million children in the US alone, and can make developing friendships + understanding social cues *extremely* difficult.
👉 Learn how @Stanford is using @TensorFlow to help children understand emotions in real-time: med.stanford.edu/news/all-news/…
👫 Autism impacts >1 million children in the US alone, and can make developing friendships + understanding social cues *extremely* difficult.
👉 Learn how @Stanford is using @TensorFlow to help children understand emotions in real-time: med.stanford.edu/news/all-news/…
⚕️ An additional note:
If you have a child who is between 3 - 17yo with a clinical diagnosis of ASD, and you live within driving distance to @Stanford: the research team would love to connect with you / have you join their study!
👉 Get involved here: autismglass.stanford.edu
If you have a child who is between 3 - 17yo with a clinical diagnosis of ASD, and you live within driving distance to @Stanford: the research team would love to connect with you / have you join their study!
👉 Get involved here: autismglass.stanford.edu
#TFadvent #10
✨🌎 This @MDPIOpenAccess paper uses @TensorFlow Hub to classify landscapes from satellites, aircraft, unmanned aerial vehicles, and fixed camera installations. Exciting impacts for geoscientists!
mdpi.com/2076-3263/8/7/…
🙏 Thanks @JeremiahHarmsen for sharing!
✨🌎 This @MDPIOpenAccess paper uses @TensorFlow Hub to classify landscapes from satellites, aircraft, unmanned aerial vehicles, and fixed camera installations. Exciting impacts for geoscientists!
mdpi.com/2076-3263/8/7/…
🙏 Thanks @JeremiahHarmsen for sharing!
✨🌎 A friend asked about implications of this research, so am sharing the answer here, too.
❤️ There are many reasons to get excited about object segmentation applied to satellite imagery: but having the ability to create *MASSIVE* new time-series geo data sets is my favorite.
❤️ There are many reasons to get excited about object segmentation applied to satellite imagery: but having the ability to create *MASSIVE* new time-series geo data sets is my favorite.
#TFadvent Day #11
@Allegorithmic announced #ProjectAlchemist at #SIGGRAPH18: a new standalone tool that combines several @TensorFlow-enabled methods for creating augmented materials.
🎬 Blog post + keynote: animationmagazine.net/technology/the…
And check out the #SubstanceAlchemist hashtag!
@Allegorithmic announced #ProjectAlchemist at #SIGGRAPH18: a new standalone tool that combines several @TensorFlow-enabled methods for creating augmented materials.
🎬 Blog post + keynote: animationmagazine.net/technology/the…
And check out the #SubstanceAlchemist hashtag!
#12!
🍞 Kasha is often contaminated with wheat grains, which are difficult to spot and can be disastrous for folks who suffer from #celiac (est. 3 million in the US alone!).
👉 Check out this @Android wheat-spotter, built with @GoogleColab + @TensorFlow: github.com/failure-to-thr…
🍞 Kasha is often contaminated with wheat grains, which are difficult to spot and can be disastrous for folks who suffer from #celiac (est. 3 million in the US alone!).
👉 Check out this @Android wheat-spotter, built with @GoogleColab + @TensorFlow: github.com/failure-to-thr…
#TFadvent #13
🔍 Have you ever wanted your own, personal "ENHANCE" button? (tvtropes.org/pmwiki/pmwiki.…)
You can get relatively close with #SRGAN: a generative adversarial network that gives pixelated images super-resolution.
👩💻: github.com/tensorlayer/sr…
📰: arxiv.org/pdf/1609.04802…
🔍 Have you ever wanted your own, personal "ENHANCE" button? (tvtropes.org/pmwiki/pmwiki.…)
You can get relatively close with #SRGAN: a generative adversarial network that gives pixelated images super-resolution.
👩💻: github.com/tensorlayer/sr…
📰: arxiv.org/pdf/1609.04802…
✨#TFadvent #14!
🧝♀️Check out this project from @farzatv where he uses @TensorFlowJS to detect which @PlayOverwatch character someone is using - then plays music specific to that character with mood.gg. 🎶
📰: medium.com/@farzatv/deepo…
👩💻: github.com/farzaa/DeepOve…
🧝♀️Check out this project from @farzatv where he uses @TensorFlowJS to detect which @PlayOverwatch character someone is using - then plays music specific to that character with mood.gg. 🎶
📰: medium.com/@farzatv/deepo…
👩💻: github.com/farzaa/DeepOve…
🛸 #TFadvent #15:
Reaver, by @Inoryy: a @StarCraft II Deep Reinforcement Learning Agent, following in @DeepMindAI's footsteps. Works with Gym, #Atari, and MuJoCo, as well!
💻 Code: github.com/Inoryy/reaver-…
📒 @GoogleColab: colab.research.google.com/drive/1DvyCUdy…
📽️ Demo:
Reaver, by @Inoryy: a @StarCraft II Deep Reinforcement Learning Agent, following in @DeepMindAI's footsteps. Works with Gym, #Atari, and MuJoCo, as well!
💻 Code: github.com/Inoryy/reaver-…
📒 @GoogleColab: colab.research.google.com/drive/1DvyCUdy…
📽️ Demo:
📚 #TFadvent #16
tf.wiki: an open-source Mandarin / English @TensorFlow textbook authored by one of our @GoogleDevs Experts (GDEs), @snowkylin.
👉 If you've been wanting to understand eager execution + tf.keras, this is a great resource!
tf.wiki: an open-source Mandarin / English @TensorFlow textbook authored by one of our @GoogleDevs Experts (GDEs), @snowkylin.
👉 If you've been wanting to understand eager execution + tf.keras, this is a great resource!
✨🔍 #TFadvent #17!
Beholder, by Chris Anderson: a @TensorBoard plugin for viewing frames of a video while your model trains. It comes with tools to visualize the parameters of your network, visualize arbitrary arrays like gradients, and more.
👉 Repo: github.com/chrisranderson…
Beholder, by Chris Anderson: a @TensorBoard plugin for viewing frames of a video while your model trains. It comes with tools to visualize the parameters of your network, visualize arbitrary arrays like gradients, and more.
👉 Repo: github.com/chrisranderson…
✨📊 Note: this is a community-built plug-in that was incorporated back into @TensorBoard itself. Thanks for your work, Chris!
If you have interest in building a plug-in like Beholder to support interpretability in #MachineLearning, check out this guide:
github.com/tensorflow/ten…
If you have interest in building a plug-in like Beholder to support interpretability in #MachineLearning, check out this guide:
github.com/tensorflow/ten…
#TFadvent #18!
TensorSpace: a 3D neural network visualization framework built with @TensorFlow.js, Three.js, and Tween.js.
📊 TensorSpace provides #Keras-like APIs to build layers, load pre-trained models, and generate browser-based 3D visualizations!
github.com/tensorspace-te…
TensorSpace: a 3D neural network visualization framework built with @TensorFlow.js, Three.js, and Tween.js.
📊 TensorSpace provides #Keras-like APIs to build layers, load pre-trained models, and generate browser-based 3D visualizations!
github.com/tensorspace-te…
#TFadvent #19!
Grasp2Vec: learning object representations through self-supervised grasping. This @GoogleAI + @UCBerkeley + XRobotics collaboration was inspired by object permanence.
Note to self: take to arcades... 😃✨
📜 arxiv.org/pdf/1811.06964…
🤖 sites.google.com/site/grasp2vec/
Grasp2Vec: learning object representations through self-supervised grasping. This @GoogleAI + @UCBerkeley + XRobotics collaboration was inspired by object permanence.
Note to self: take to arcades... 😃✨
📜 arxiv.org/pdf/1811.06964…
🤖 sites.google.com/site/grasp2vec/
#TFadvent #20!
🗺️ Take a Look Around: using @Google #StreetView, @TensorFlow + #Keras, and satellite imagery to estimate housing prices from the "visual appeal" of neighborhoods, via researchers from the @turingist, @UCL, @Cambridge_Uni.
📜 arxiv.org/pdf/1807.07155…
🗺️ Take a Look Around: using @Google #StreetView, @TensorFlow + #Keras, and satellite imagery to estimate housing prices from the "visual appeal" of neighborhoods, via researchers from the @turingist, @UCL, @Cambridge_Uni.
📜 arxiv.org/pdf/1807.07155…
#TFadvent #21
☀️ DeepSolar is a deep learning project from @Stanford that analyzes satellite imagery to identify the GPS locations and sizes of solar photovoltaic (PV) panels.
👉 For more information, including a link to their @TensorFlow code on GitHub: web.stanford.edu/group/deepsola…
☀️ DeepSolar is a deep learning project from @Stanford that analyzes satellite imagery to identify the GPS locations and sizes of solar photovoltaic (PV) panels.
👉 For more information, including a link to their @TensorFlow code on GitHub: web.stanford.edu/group/deepsola…
#TFadvent #22
🐢 Don't be a Turtle: real-time posture correction for programmers using @TensorFlow, @Android, and #TFlite! Making a mental note to download this... 😁
Created by one of our ML @GoogleDevExpert members! 🙏
👩💻 Check out his @github here: github.com/motlabs/dont-b…
🐢 Don't be a Turtle: real-time posture correction for programmers using @TensorFlow, @Android, and #TFlite! Making a mental note to download this... 😁
Created by one of our ML @GoogleDevExpert members! 🙏
👩💻 Check out his @github here: github.com/motlabs/dont-b…
#TFadvent #23
BigGAN: a new state-of-the-art for image synthesis, improving Inception scores by > 100. 😮
Created by Andrew Brock, an intern (!!) from @HeriotWattUni, in collaboration with researchers from @DeepMindAI. Talk about a summer project! 💪
📄 arxiv.org/abs/1809.11096
BigGAN: a new state-of-the-art for image synthesis, improving Inception scores by > 100. 😮
Created by Andrew Brock, an intern (!!) from @HeriotWattUni, in collaboration with researchers from @DeepMindAI. Talk about a summer project! 💪
📄 arxiv.org/abs/1809.11096
#TFadvent #24!
🎨 One of our ML @GoogleDevExpert team told me about this @TensorFlow project that dates ancient, historic paintings in the Mogao grottoes.
3860 mural paintings collected from 194 locations in northwestern China. So beautiful! 😍✨
📄: link.springer.com/article/10.100…
🎨 One of our ML @GoogleDevExpert team told me about this @TensorFlow project that dates ancient, historic paintings in the Mogao grottoes.
3860 mural paintings collected from 194 locations in northwestern China. So beautiful! 😍✨
📄: link.springer.com/article/10.100…
🎅 #TFadvent #25!
From @pyimagesearch, and just in time for #ChristmasMorning: building a Santa / not Santa detector with @TensorFlow + #Keras and a @Raspberry_Pi.
🤖 Would love to see this project updated to use #TFLite!
👉 Blog post with instructions: pyimagesearch.com/2017/12/18/ker…
From @pyimagesearch, and just in time for #ChristmasMorning: building a Santa / not Santa detector with @TensorFlow + #Keras and a @Raspberry_Pi.
🤖 Would love to see this project updated to use #TFLite!
👉 Blog post with instructions: pyimagesearch.com/2017/12/18/ker…
