Book a Google I/O 2024 Conference Live Now! Encyclopedia Explains KAN, a New Neural Network Architecture That Became Popular Overnight
The Google I/O 2024 Keynote will be held at 1 a.m. on May 15 (next Wednesday).HyperAI will broadcast live on Video Account and Bilibili.
After the live broadcast, there will be a "Too Long to Read" version of the meeting highlights for everyone!Remember to follow the official account to get first-hand information!
From May 6 to May 10, hyper.ai official website updates:
* High-quality public datasets: 10
* Selection of high-quality tutorials: 3
* Community Article Selection: 6 articles
* Popular encyclopedia entries: 5
Visit the official website:hyper.ai
Selected public datasets
1. MUSTARDSAUCE Mathematical Theorem Problem Dataset
The data of MUSTARDSAUCE includes math word problems and theorem proof problems, covering the difficulty level from primary school to higher education. The number of reasoning steps of the questions increases with the difficulty of the questions, and the most difficult questions require about 30 steps to solve.
Direct use:https://go.hyper.ai/RyOhj
2. VisA Industrial Visual Anomaly Detection Dataset
The VisA dataset is a SPot-the-Difference self-supervised pre-training dataset for anomaly detection and segmentation. It contains 12 subsets corresponding to 12 different objects. There are 10,821 images, including 9,621 normal samples and 1,200 abnormal samples.
Direct use:https://go.hyper.ai/IdZr5
3. Michelin Guide Restaurants Michelin restaurant dataset
This dataset contains a list of restaurants and other detailed information (such as address, price range, cuisine type, longitude, latitude, etc.) compiled according to the Michelin restaurant guide. Specifically, it includes Michelin 3-star, 2-star, 1-star, Bib Gourmand, and Green Star restaurants. This restaurant list will show the address (Address), price range (Price), cuisine type (Cuisine), and other information of Michelin restaurants.
Direct use: https://go.hyper.ai/HNQAy
4. European 515K Hotel Review Dataset
This dataset contains more than 510,000 customer reviews and ratings of 1,493 luxury hotels in Europe, and also provides the hotel's location for further analysis. The data was scraped from the Booking website.
Direct use: https://go.hyper.ai/gwsIs
5. Geolocation-food Dish Food Image Dataset
The dataset contains 117,504 dishes from 6 regions. The related paper was published in IEEE in 2015.
Direct use: https://go.hyper.ai/jxujh
6. MPDD Metal Parts Defect Detection Dataset
This dataset is a metal parts defect visual detection dataset, suitable for anomaly detection tasks in images in industrial scenarios.
Direct use: https://go.hyper.ai/iMyTO
7. Stargazer 88 constellation image dataset
This dataset consists of hand-drawn images of all 88 constellation types and their annotations in JSON format, including two types of files: train and test.
Direct use: https://go.hyper.ai/F12fI
8. SCUT-FBP500 V2 Facial Beauty Rating Dataset
The dataset contains 5,500 images covering Asian and Caucasian, male and female faces. Each face is rated on a scale of 1 to 5, providing valuable data for cross-cultural beauty standards and personal preference tasks.
Direct use:https://go.hyper.ai/1Jc2k
9. Celebrity Face 18 celebrity face image dataset
This dataset contains images of 18 Hollywood celebrities, with 100 images for each celebrity, for a total of 1,800+ images.
Direct use:https://go.hyper.ai/ls5AN
10. Celebrity Faces 50K celebrity face image dataset
This dataset is a curated subset of the CelebFaces Attributes (CelebA) dataset, carefully selected for deep learning tasks such as image synthesis and face recognition. It contains 50,000 images of celebrity faces of different identities, covering a wide range of poses, backgrounds, and facial attributes. These images are suitable for experimenting with GANs, face recognition models, and other machine learning tasks related to face analysis.
Direct use:https://go.hyper.ai/kLgFZ
For more public datasets, please visit:
Selected Public Tutorials
APISR is an open source project dedicated to improving resolution. It can not only restore and enhance low-quality, low-resolution anime images and video sources, but also handle various image degradation problems (such as blur, noise, compression artifacts, etc.) and provide flexible upscaling options.
Run online:https://go.hyper.ai/8msXE
2. SegmentAnything source code implementation and online reasoning
Segment Anything Model (SAM) is a machine vision model that can generate high-quality image segmentation based on input prompts such as points or boxes, and can be used to generate corresponding masks for all objects in an image. The model is trained on a dataset of 11 million images and 1.1 billion masks, and has strong zero-shot performance on various segmentation tasks, achieving true segmentation of everything.
Run online:https://go.hyper.ai/NT2mJ
3. InstantID personalized photo generation demo
InstantID is an image generation solution based on a diffusion model that can quickly generate a variety of stylized portraits from a single reference image. Just upload a selfie and get a customized AI portrait in 20 seconds.
Run online:https://go.hyper.ai/Eq4zl
Community Articles
With the continuous development of computer vision technology, image segmentation has shown important application value in many fields. This week, HyperAI has compiled and analyzed 10 high-quality image segmentation datasets to better promote the progress of related research.
Get detailed information:https://go.hyper.ai/gg6W7
This article starts with the concept, introduces the mainstream super-resolution algorithms and their advantages and disadvantages, and shows its specific application cases in different fields. It also summarizes public tutorials and public datasets related to super-resolution.
View the full report:https://go.hyper.ai/NfXvO
On May 8th, local time, Google DeepMind and its subsidiary Isomorphic Labs released AlphaFold 3. This model once again refreshed people's cognition. How did it achieve "all molecules of life are predictable"? What breakthroughs have been made in technology and performance? What role does Isomorphic Labs, which was born out of DeepMind, play? This article will give you the answers!
View the full article:https://go.hyper.ai/3uBMM
Stanford University's Human-Center Artificial Intelligence Research Center released the "2024 Artificial Intelligence Index Report". The report shows the research results of AI in 2023 in physical simulation, chemistry, materials, weather forecasting, flood forecasting, etc. This article is a detailed interpretation and sharing of the AI for Science content in the report.
View the full report:https://go.hyper.ai/2cdT5
5. Google releases HEAL framework, 4 steps to evaluate whether medical AI tools are fair
The Google team developed the HEAL framework, which can quantitatively evaluate whether medical solutions based on machine learning are fair. The team tried to ensure that emerging health technologies can effectively reduce health inequalities rather than exacerbate them. This article is a detailed interpretation and sharing of the research.
View the full report:https://go.hyper.ai/xlEa7
The brain science research team of Fudan University borrowed the idea of "semantic segmentation" from computer vision, proposed the concept of "spatial transcriptome semantic annotation", and developed the spatial transcriptome semantic annotation tool Pianno, which can enhance the interpretation of complex biological systems. This article is a detailed interpretation and sharing of the research.
View the full report:https://go.hyper.ai/FTZeo
Popular Encyclopedia Articles
1. Epoch
2. Reciprocal ranking fusion RRF
3. Masked Language Modeling (MLM)
4. Kolmogorov-Arnold Representation Theorem
5. Kolmogorov-Arnold Networks (KAN)
Here are hundreds of AI-related terms compiled to help you understand "artificial intelligence" here:
Station B live broadcast preview
Google will hold Google I/O 2024 on May 14th local time. HyperAI Super Neural Video Account and Bilibili will broadcast it live. Please scan the QR code below to make an appointment immediately!
To help you gain a deeper understanding of Google,The Chao Shen Jing live broadcast room on Bilibili will continue to broadcast the "Google Special" video.Involves: Google I/O press conferences over the years, interviews with executives, related documentaries and other rich content.
The following table is a preview of the content selected by the editor↓↓↓
| date | time | content |
| May 13 Monday | 18:00 | Google I/O Conferences over the Years |
| Tuesday, May 14 | 18:00 | Google Cloud NEXT Conferences |
| May 15Wednesday | 1:00 | Google I/O 2024 Keynote |
| Thursday, May 16 | 18:00 | Google CEO on the US-China AI race |
| Friday, May 17 | 18:00 | AlphaGo Documentary |
| Saturday, May 18 | 18:00 | TIME100 Interview with Sundar Pichai |
| Sunday, May 19 | 18:00 | BBC documentary: A World Without Google |
Super Neuro TV broadcasts live 24/7. Click to get the "electronic pickles" in the AI field:
http://live.bilibili.com/26483094
The above is all the content of this week’s editor’s selection. If you have resources that you want to include on the hyper.ai official website, you are also welcome to leave a message or submit an article to tell us!
See you next week!
About HyperAI
HyperAI (hyper.ai) is the leading artificial intelligence and high-performance computing community in China.We are committed to becoming the infrastructure in the field of data science in China and providing rich and high-quality public resources for domestic developers. So far, we have:
* Provide domestic accelerated download nodes for 1200+ public data sets
* Includes 300+ classic and popular online tutorials
* Interpretation of 100+ AI4Science paper cases
* Support 500+ related terms search
* Hosting the first complete Apache TVM Chinese documentation in China
Visit the official website to start your learning journey: