GIFT: Learning Transformation-Invariant Dense Visual Descriptors via Group CNNs https://t.co/dmx0NTxo4U #computervision #NeurIPS2019 pic.twitter.com/CDqw1dsg8o
— Tomasz Malisiewicz (@quantombone) November 15, 2019
GIFT: Learning Transformation-Invariant Dense Visual Descriptors via Group CNNs https://t.co/dmx0NTxo4U #computervision #NeurIPS2019 pic.twitter.com/CDqw1dsg8o
— Tomasz Malisiewicz (@quantombone) November 15, 2019
We are pleased to announce the release of source code and checkpoints for MobileNetV3 and the Pixel 4 Edge TPU-optimized counterpart, MobileNetEdgeTPU, advancing state-of-the-art for computer vision tasks on compute and power-limited devices. Learn more ↓ https://t.co/qPOKiJOFP4
— Google AI (@GoogleAI) November 13, 2019
They iteratively train several models. Largest is trained for 3.5 days on 2048-v3 TPU pod. It seems that one hour of this pod is ~$2.5K. Thus, we get an insane price tag ~$200K. The full experiment might cost you about half a MILLION USD if u want to reproduce it. https://t.co/wthQXaGe40
— Leonid Boytsov (@srchvrs) November 13, 2019
Nice new results from @GoogleAI researchers on improving the state-of-the-art on ImageNet!
— Jeff Dean (@JeffDean) November 12, 2019
"We...train a...model on...ImageNet...& use it as a teacher to generate pseudo labels on 300M unlabeled images. We then train a larger...model on the...labeled & pseudo labeled images." https://t.co/KGbKRkQt07
torchvision v0.4.2: Optimized video reader backend
— PyTorch (@PyTorch) November 7, 2019
This minor release provides up to 6x speedup for video reading using a new backend option.
Read more at: https://t.co/p2nKoKUX5g
We’re pleased to release the Visual Task Adaptation Benchmark (VTAB), a diverse, realistic, and challenging protocol to measure progress towards universal visual representations. Learn all about it below. https://t.co/PbORwSFPAg
— Google AI (@GoogleAI) November 6, 2019
UR-FUNNY: A Multimodal Language Dataset for Understanding Humor
— hardmaru 😷 (@hardmaru) November 5, 2019
They used TED Talk transcripts with laughter cues to create a humor dataset that can be used for humor detection and other humor analyses. #EMNLP2019 🤣
paper https://t.co/HGYDViRRoE
dataset https://t.co/MedSrcf07s pic.twitter.com/kTplhlK5oT
New tutorial!🚀 Traffic Sign Classification with #Keras and #TensorFlow 2.0 🛑🚦⚠️
— Adrian Rosebrock (@PyImageSearch) November 4, 2019
- 95% accurate
- Includes pre-trained model
- Full tutorial w/ #Python codehttps://t.co/MkWiTaYKwU 👍#DeepLearning #MachineLearning #ArtificialIntelligence #DataScience #AI pic.twitter.com/20jYvQw7Dw
Pretrained EfficientNet, MixNet, MobileNetV3, MNASNet A1 and B1, FBNet, Single-Path NAS https://t.co/1cCSJGrNrN #deeplearning #machinelearning #ml #ai #neuralnetworks #datascience #pytorch
— PyTorch Best Practices (@PyTorchPractice) November 3, 2019
Facebook highlights AI that converts 2D objects into 3D shapes - Easy read for people interested in an overview of these computer vision efforts. https://t.co/JNCaZ52svL
— Nando de Freitas (@NandoDF) October 30, 2019
Winning entry of the 2019 video activity detection challenge now open source. https://t.co/x3PrTcRyV6
— Yann LeCun (@ylecun) October 29, 2019
auto-tinder - 🖖 Train an artificial intelligence to play tinder for you https://t.co/HFWJdOAdBS
— Python Trending (@pythontrending) October 29, 2019