Content Overview: Following DALL-E and ChatGPT, OpenAI has made another effort and recently released Point·E, which can directly generate 3D point clouds based on text prompts. Keywords: OpenAI 3D point cloud Point·E OpenAI is rushing to achieve performance at the end of the year, […]

Content at a glance: While deep synthesis services meet user needs and improve user experience, they are also used by some lawless persons to produce, copy, publish, and disseminate illegal information, defame and denigrate the reputation and honor of others, and impersonate others to commit fraud and other illegal acts. Now the management regulations for this technology have finally been released. Keywords: De […]

The purpose of NAS is to discover the best architecture for a neural network. The integration of Torch and Ax provides assistance for multi-objective exploration of neural architectures. This article will show how to run a fully automatic neural architecture search using Multi-objective Bayesian NAS. […]

PyTorch 2.0 can improve model training speed by 30%-200% with a simple line of torch.compile(). This tutorial will show how to actually reproduce this speedup. torch.compile() makes it easy to try different compilers […]

Contents at a glance: This section explains how to use TVMC to compile and optimize models. TVMC is the command driver for TVM, which executes TVM functions through command lines. This section is the basis for understanding how TVM works. Keywords: TVMC TVM& […]

Contents at a glance: There are three ways to install TVM: install from source code, install using Docker image, and install using NNPACK Contrib. This article explains how to install using Docker image and NNPACK Contrib. Keywords: TVM […]

Contents at a glance: YOLO v4 is a real-time, high-precision target detection model. This tutorial will explain in detail how to implement crowd distance detection under multi-target conditions based on YOLO v4 and SORT algorithms. Keywords: YOLO v4 SORT Multi-target detection The outbreak of the new crown epidemic began at the beginning of the […]

Contents at a glance: At the PyTorch Conference 2022 held last night, PyTorch 2.0 was officially released. This article will sort out the biggest differences between PyTorch 2.0 and 1.x. Keywords: PyTorch 2.0 compiler […]

In the previous article, we introduced TorchMultimodal. Today, we will start from a specific case to demonstrate how to expand the multimodal basic model in the TorchMultimodal library with the support of Torch Distributed technology. In recent years […]

Contents at a glance: There are three ways to install TVM: installing from source code, installing using Docker images, and installing using NNPACK Contrib. This article focuses on how to install TVM from source code. Keywords: TVM Quick Start Source Code Installation In the previous article "TVM […]

The graph mode in PyTorch is more powerful in terms of performance. This article introduces Torch.FX, a powerful tool that can capture and optimize the graph of PyTorch programs. 1. Introduction PyTorch supports two execution modes: eager […]

​ Content Overview: On November 22, in the first match of the World Cup Group C, Saudi Arabia overtook Argentina 2:1. Today, we will review the technology and hard work behind Argentina's 4 goals and 1 point. Keywords: World Cup VAR semi-automatic offside technology This article was first published on WeChat official account: Hy […]

Multimodal artificial intelligence is a new AI paradigm that combines multiple data types such as images, text, voice, and video with multiple intelligent processing algorithms to achieve higher performance. Recently, PyTorch officially released a domain library-TorchM […]

150+ Chinese documents, searchable

Contents at a glance: One of the types of Chinese text errors is spelling errors. This article is a model deployment tutorial for using the BART pre-training method to implement Chinese text error correction. Keywords: BART Chinese spelling correction NLP This article was first published on WeChat official account: HyperAI Super Neural Chinese Text Error 3 […]

Contents at a glance: This issue summarizes 9 datasets recently updated on the HyperAI official website, covering three areas: face recognition, posture estimation, and autonomous driving. Keywords: face recognition, posture estimation, autonomous driving This article was first published on WeChat official account: HyperAI Recently, HyperAI […]

Content introduction: TorchVision Transforms API has been expanded and upgraded to support object detection, instance and semantic segmentation, and video tasks. The new API is still in the beta stage and developers can try it out. This article was first published on WeChat official account: PyTorch Developer Community […]

Contents at a glance: Container number detection based on PaddleOCR shortens the time of recording container numbers and improves port loading and unloading efficiency. Keywords: PaddleOCR text recognition online tutorial International shipping consulting and analysis agency Alphaliner released a set of data in March this year […]

Gama Yuanwang Analysis Platform is committed to providing visual and intelligent data insight solutions for enterprises, banks and other non-bank financial institutions.

In today's data-driven world, every subtle fluctuation in data may contain unlimited business opportunities. However, faced with the integration, management and application of massive data, enterprises often feel overwhelmed. How to break the data silos, make data a real asset, and promote the digital transformation of enterprises? Today, we will reveal to you […]

In this digital age, every step of innovation in the financial industry is crucial. Faced with an increasingly complex risk environment and increasing customer experience requirements, how can financial institutions provide efficient and intelligent services while ensuring security? Today, let us explore the intelligent face-to-face review solution of the OneConnect Gama platform and explore […]

AnimeGANv2 recently released an update developed by community contributors, and a demo that can be run online was implemented through Gradio and published on huggingface. Visit https://huggingface.co/spaces/ […]

Last month, the well-known Taiwanese YouTube celebrity "Xiao Yu" was formally arrested on charges of distributing obscene materials and defamation and was transferred to the New Taipei District Prosecutor's Office for investigation. The police announced that Xiao Yu and other suspects used DeepFake's AI face-changing technology to recruit more than 8,000 members and illegally obtained […]

Archaeology has always been a mysterious and attractive topic. In the tomb-robbing novel series "Ghost Blowing the Light", the protagonist Hu Bayi relied on half of the "Sixteen-Character Yin-Yang Feng Shui Secret Art" to master the art of finding dragons and dividing gold to locate the tomb, and found the big tomb and treasure. Hu Bayi's formula: Finding dragons and dividing gold depends on the winding mountains. Each winding is a barrier. If the gate is eight layers of danger, it is not […]

After bidding farewell to the last statutory holiday in 2021, everyone returned to work and the rhythm of commuting. On the map navigation software of various large office areas in Beijing, Shanghai, Guangzhou and Shenzhen, the lines also turned back to yellow, dark red, and even purple during rush hour. Google released an article titled […]

The DeepSOCIAL crowd distance monitoring project, which was open-sourced by Mahdi Rezaei, a researcher at the Institute of Transport Studies at the University of Leeds, has become a popular online application recently. Online source code tutorial access DeepSOCIAL uses YOLOv4 to achieve pedestrian detection, and then uses SORT […]

On a private farm in Ohio, USA, Shay Myers owns 30 acres (about 120 mu) of onion planting land. Three generations of his family have been engaged in agricultural production and have experienced the development of agricultural mechanization in the United States for nearly a hundred years. This year, Shay Myers used two 3-meter-long, 4-ton large […]

Since April 2020, the Seoul National University of Science and Technology in South Korea has been working with Seoul’s emergency services to analyze and predict whether pedestrians are suicidal based on surveillance footage from the Han River Bridge. Once AI predicts a dangerous situation or suspicious behavior, it will immediately alert the rescue team, who will […]

​ Recently, a movie based on video games, Free Guy (Chinese translation: Out of Control Player), was released in North America and China, starring Ryan Reynolds, the actor of Deadpool. The whole movie is lighthearted and funny, with exquisite special effects. It performed well in the summer season and gained good box office results and many audiences around the world. […]