Hey! Got 40 seconds? ⏱️ Learn how we achieve photorealistic reposing and virtual try-on in the upcoming SIGGRAPH Asia paper *Pose with Style*. 🤩
— Jia-Bin Huang (@jbhuang0604) October 15, 2021
Paper: https://t.co/UtVyBn8eA3
Web: https://t.co/b55nH3SSDB
Brought to you by the amazing @BadourAlBahar! pic.twitter.com/Xf9jYFaOY5
ByteTrack: Multi-Object Tracking by Associating Every Detection Box
— AK (@ak92501) October 14, 2021
abs: https://t.co/x2es6mhWh4
github: https://t.co/AuNaKamE9s@XinggangWang pic.twitter.com/dsVm5TzVb0
Object-Region Video Transformers
— AK (@ak92501) October 14, 2021
abs: https://t.co/X0dS3HSYaT
project page: https://t.co/yDAErxp4zE
an object-centric approach that extends video transformer layers with a block that directly incorporates object representations pic.twitter.com/b7ozNo7I0B
Medical image classification models often pre-train on natural image datasets. Today, we present alternative approaches that use additional pre-training on medical images, along with metadata-based data augmentation, to significantly improve performance. https://t.co/DHy0XojZwm
— Google AI (@GoogleAI) October 13, 2021
Data-free Knowledge Distillation for Object Detection
— AK (@ak92501) October 13, 2021
pdf: https://t.co/y6ZqK9jenm
github: https://t.co/ncFsFtIsTX pic.twitter.com/e7OwRU9F7C
Distilling Image Classifiers in Object Detectors
— AK (@ak92501) October 13, 2021
abs: https://t.co/yVJW5hQOk5
github: https://t.co/8JVvrROA1f pic.twitter.com/9iSqm17Upo
TAda! Temporally-Adaptive Convolutions for Video Understanding
— AK (@ak92501) October 13, 2021
abs: https://t.co/uv6Jqnu580 pic.twitter.com/Suh5Te7P67
Causal ImageNet: How to discover spurious features in Deep Learning?
— AK (@ak92501) October 12, 2021
abs: https://t.co/NvelZA7Aa4 pic.twitter.com/OhFgarP17L
ViDT: An Efficient and Effective Fully Transformer-based Object Detector
— AK (@ak92501) October 11, 2021
abs: https://t.co/rOytM75swG
obtains the best AP and latency trade-off among existing fully transformer-based object detectors, and achieves 49.2AP owing to its high scalability for large models pic.twitter.com/CVAzoT3dNh
Token Pooling in Visual Transformers
— AK (@ak92501) October 11, 2021
abs: https://t.co/0Jr3cJqvRe
Applied to DeiT, achieves the same ImageNet top-1 accuracy using 42% fewer computations pic.twitter.com/lusPaB9Bns
We’re using KaoKore, a machine learning-friendly dataset, to decipher cursive and illustrations in historical Japanese art. Learn how machine learning can be used for humanities research and contribute to cultural preservation. https://t.co/nYSvuehWoR
— Boo-gle 👻 (@Google) September 30, 2021
Amusing! Object detection cast naively into language modeling framework + borrowing many of the tips&tricks.
— Andrej Karpathy (@karpathy) September 24, 2021
- random object ordering seems fine ✅
- coords, class labels flattened into a single softmax 😂
- sequence augmentation is the most gnarly part, almost as yucky as nms 😬 https://t.co/FxSz5UbpxY
