Tencent Launches PhotoMaker, a New Level of Personalized Image Generation; LawInstruct, the First Large-scale Legal Instruction Dataset, Is Launched
PhotoMaker is an AI image generation tool jointly developed by Tencent and Nankai University. It can not only capture the facial features of people, but also generate realistic images of people according to their preferences, pushing personalization to a new level. Whether you want a brand new social media avatar or design a unique character for a game, it can satisfy you.
The one-click run demo tutorial of this project is now available on the hyper.ai official websiteIn order to make it easier and quicker for everyone to experience the model, the ComfyUI workflow version is also available, which is faster to generate! More custom forms are available! Come and experience it~
From June 17 to June 21, hyper.ai official website updates:
* High-quality public datasets: 10
* Selected high-quality tutorials: 2
* Community article selection: 4 articles
* Popular encyclopedia entries: 5
* Top conferences with deadlines in June and July: 5
Visit the official website:hyper.ai
Selected public datasets
1. TreeOfLife-10M biological image dataset
The dataset has more than 10 million images covering 454,000 taxa in the tree of life, making it the largest dataset of biological organism images and their associated classification labels for ML to date. The research paper was selected as the best paper of CVPR 2024.
Direct use:https://go.hyper.ai/jimyx
2. HAR 15 human action recognition dataset
The dataset is a base containing 15 different categories of human activities, about 12k+ labeled images (including verification images), each image has only one human activity category, and aims to understand human behavior and assign labels to each action.
Direct use:https://go.hyper.ai/3jmf1
3. Linux China Original Articles Dataset
This dataset is a dataset of Linux China original technical articles + translated technical articles, providing multiple fields such as article title and content.
Direct use:https://go.hyper.ai/KgFgf
4. DroneVehicle Large-scale drone aerial vehicle detection dataset
The dataset consists of 56,878 images collected by drones, half of which are RGB images and the rest are infrared images. This dataset provides support for smart city traffic management and disaster relief.
Direct use:https://go.hyper.ai/KhbuB
5. PianoMotion10M Piano Motion Gesture Generation Dataset
This dataset is a benchmark dataset focused on piano playing gesture generation, covering 116 hours of piano playing videos shot from a bird's-eye view and containing 10 million annotated hand gestures. These data provide a rich resource to explore the relationship between piano music and hand movements.
Direct use:https://go.hyper.ai/QAOi8
6. LawInstruct: The first large-scale legal instruction dataset
This dataset is the first large-scale instruction dataset for the legal field, covering 17 jurisdictions and 24 languages, containing 12 million training examples, and covering a variety of legal tasks such as question answering, entailment, summarization, and information extraction.
Direct use:https://go.hyper.ai/ARN3n
7. ASL Alphabet Sign Language Dataset
This dataset contains images of American Sign Language letters, which are divided into 29 categories, representing different sign language letters and symbols. The image size in the dataset is unified to 200×200 pixels, and the training dataset contains 87,000 images.
Direct use:https://go.hyper.ai/IhhZc
8. AZ handwritten letter dataset in .csv format
This dataset is a large-scale collection of English handwritten letter images, designed specifically for handwriting recognition tasks and can be used to train various image recognition models.
Direct use:https://go.hyper.ai/28lwk
9. CAT dataset Cat facial feature annotation image dataset
This dataset contains over 9,000 cat images annotated with facial features. Each image has an annotation of a cat head with 9 points: 1 for the mouth, 2 for the eyes, and 6 for the ears.
Direct use:https://go.hyper.ai/MkmTu
The dataset contains hundreds of audio files of cats and dogs: 164 WAV files for cats, corresponding to 1,323 seconds of audio; 113 WAV files for dogs, corresponding to 598 seconds of audio. All WAV files contain 16KHz audio and are of variable length.
Direct use:https://go.hyper.ai/e0RGg
For more public datasets, please visit:
Selected Public Tutorials
1. PhotoMaker arbitrary stylized image generation demo
PhotoMaker is an efficient portrait customization model that can quickly generate customized portrait photos and artistic photos based on portrait photos. In addition to generating personalized portrait photos, it can also change the age and gender of the person, integrate the characteristics of different people to create new person information, and is a very useful AI painting tool. This tutorial has set up the environment for everyone to run and experience.
Run online:https://go.hyper.ai/3B5Rd
2. ComfyUI PhotoMaker ZHO image generation workflow
This tutorial is about how to run the ComfyUI workflow in PhotoMaker. Compared with the Gradio version of the Demo, the ComfyUI workflow version has the advantages of faster generation, support for node splitting and custom size, etc. Chao Neural Network has built the relevant environment for you, and you can run it with one-click cloning.
Run online:https://go.hyper.ai/XdSoQ
3. Song style classification based on HuBERT
This tutorial uses the base model HuBERT to demonstrate how to use a pre-trained audio model to perform audio classification tasks, and implements music style classification by fine-tuning on the pre-trained model.
Run online:https://go.hyper.ai/7QjRj
We have also established a Stable Diffusion tutorial exchange group. Welcome friends to scan the QR code and remark [SD tutorial] to join the group to discuss various technical issues and share application results~
Community Articles
HyperAI will hold the 5th offline gathering of Meet AI Compiler Technology Salon at the Institute of Computing Technology, Chinese Academy of Sciences on July 6 (Saturday). This meetup is divided into two parts: technology sharing and roundtable discussion. Guests are from Shanghai Jiao Tong University, Institute of Computing Technology, Chinese Academy of Sciences, Microsoft Research Asia, and Beijing Zhiyuan Artificial Intelligence Research Institute. We hope that the new gathering will bring new technical gains to everyone and make new friends in the same field.Click "Read original text" to sign up and participate!
View event details:https://go.hyper.ai/oAsLQ
The Shanghai Institute of Artificial Intelligence and Finance, the world's first education and research institution built around the intersection of artificial intelligence and finance, was officially established recently. HyperAI had the honor of having an in-depth interview with Dean Shao Yilei to further understand the vision and development plan of the new institute.
Read the full interview:https://go.hyper.ai/bp4Po
Professor Wu Sijin from the Huihu School of Pharmacy at Xi'an Jiaotong-Liverpool University, together with Professor Xie Songbo and Professor Zhong Diansheng's team from Tianjin Medical University General Hospital, developed a selective p53-R175H degrader, dp53m, which can specifically recognize the mutant p53-R175H protein and target its degradation, with significant anti-tumor efficacy and no obvious toxicity. dp53m can also enhance the sensitivity of cancer cells to the chemotherapy drug cisplatin, which is crucial for cancer treatment. This article is a detailed interpretation and sharing of the research.
View the full report:https://go.hyper.ai/k80er
Researchers from the GIS Laboratory of Zhejiang University introduced a neural network method to nonlinearly couple multiple spatial proximity measures between observation points to obtain an optimized spatial proximity measure, which improved the accuracy of the model's prediction of housing prices. The study also combined OSP with the geographic neural network weighted regression method to construct the osp-GNNWR model, which trains the neural network by solving the spatial non-stationary regression relationship between the dependent variable and the independent variable. This article is a detailed interpretation and sharing of the study.
View the full report:https://go.hyper.ai/kWmHC
Popular Encyclopedia Articles
1. Function call Function Calling
2. Spiking Neural Network (SNN)
3. Context Position Encoding CoPE
4. Russian dolls represent learning MRL
5. Kolmogorov-Arnold Representation Theorem
Here are hundreds of AI-related terms compiled to help you understand "artificial intelligence" here:
Station B live broadcast preview
Jeff Dean is a senior researcher and computer scientist at Google, known for his pioneering work in distributed systems and artificial intelligence, including the development of MapReduce and TensorFlow. He is one of the key figures in Google's technological development. This week, Super Neural TV will broadcast Jeff Dean's speeches and interviews.
The following table is a preview of the content selected by the editor↓↓↓
| date | time | content |
| Monday, June 24 | 18:00 | Jeff Dean on the five major trends in machine learning |
| Tuesday, June 25 | 18:00 | Let AI serve everyone |
| Wednesday, June 26 | 18:00 | Jeff Dean's positive outlook on the future of AI |
| Thursday, June 27 | 18:00 | Jeff Dean's speech at the Stanford Medical Big Data Conference |
| Friday, June 28 | 18:00 | Jeff Dean's talk on deep learning |
| Saturday, June 29 | 18:00 | Google Brain & Brain Residency |
| Sunday, June 20 | 18:00 | How to use deep learning to solve problems |
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 1300+ public data sets
* Includes 400+ 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: