Recently, researchers from ByteDance Research Institute and Tsinghua University jointly released a new study, pointing out that current AI video generation models, such as OpenAI's Sora, while capable of creating stunning visual effects, have significant flaws in understanding basic physical laws. This study has sparked widespread discussion on the capabilities of AI in simulating reality. The research team tested AI video generation models under three different scenarios: predictions under known patterns and predictions under unknown patterns.

Today, the Zhipu Technology team released its latest video generation model, CogVideoX v1.5, and made it open source. This version is another significant advancement in the CogVideoX series launched by the Zhipu Technology team since August. The update greatly enhances video generation capabilities, including support for 5-second and 10-second video lengths, 768P resolution, and the ability to generate 16 frames. Additionally, the I2V (Image-to-Video) model supports arbitrary aspect ratios, further improving the understanding of complex semantics.

Kuaishou, Peking University, and Beijing University of Posts and Telecommunications have joined forces to unveil a groundbreaking technological achievement—the Pyramid-Flow high-definition video model. This open-source project has made significant breakthroughs in the field of AI-generated video, bringing new possibilities to the industry. The Pyramid-Flow model demonstrates remarkable capabilities, generating high-quality videos of up to 10 seconds in length, with a resolution of 1280x768 and a frame rate of 24fps, using only text input. It excels in light and shadow effects, motion continuity, overall image quality, and text interpretation.

Recently, OpenAI has caused a stir in the video AI field as they are set to make significant upgrades to Sora, which was released in February of this year. According to reports from the Information Daily, the main goal of this upgrade is to significantly enhance Sora's performance, enabling it to generate longer and higher quality video content more quickly. Reflecting on the initial performance of Sora, there were indeed many pressing issues that needed addressing. The original version took over 10 minutes to generate a video, an efficiency level that clearly fails to meet practical application needs.

Recently, AI startup Runway announced a collaboration with renowned film production company Lionsgate. This news has caused quite a stir in the film industry, as it marks the first agreement between a major film company and an AI video model provider. Runway is headquartered in New York City and has received support from companies like Google. Although Runway's technology has already been used by some filmmakers in Hollywood, such as in the Oscar-winning film 'Everything Everywhere All at Once',

Adobe has introduced a new AI creation tool, Firefly Video Model, designed to generate high-quality video content. This tool is based on ethical training, using the company's proprietary or licensed data, and supports text-to-video generation, image-to-video conversion, and video editing, with each generated segment lasting up to 5 seconds. The videos generated during the demonstration were impressive, following text prompts and showcasing scenes from the natural world. Features include filling in video gaps, creating seamless transitions, and achieving perfect perspectives. Through Firefly, Adobe is empowering creators.

On September 4, 2024, the Beijing ZhiYuan Artificial Intelligence Research Institute (BAAI) announced the launch of the world's first model combat evaluation service featuring text-to-video, called FlagEval Large Model Arena. This service is open to users and covers approximately 40 large models from both domestic and international sources, supporting customizable online or offline evaluations for four major tasks: language Q&A, multimodal image and text understanding, text-to-image, and text-to-video.

Hedra has successfully raised $10 million in seed funding, founded by former employees from NVIDIA, Google, and the Metaverse, attracting renowned investors such as Index Ventures, Abstract, and A16Z Speedrun. The foundational model Character-1, launched by Hedra during the seed round, claims to be the fastest video model on the market, with 350,000 users having already created over 1.6 million videos, some of which have gained significant popularity on social media. The platform allows users to create...

Baidu Intelligent Cloud and Shengshu Technology have officially launched the video model Vidu, bringing new possibilities to the creative field. Vidu is the first fully self-researched video model in China, with performance fully comparable to OpenAI's Sora. It is aimed at creative fields such as art design, game production, film and animation, and social entertainment, and is dedicated to stimulating the productivity and creativity of all humanity.

In the world of AI, making machines understand videos is much harder than understanding images. Videos are dynamic, with sound, movement, and a myriad of complex scenes. In the past, AI viewed videos as if they were reading ancient scrolls, often leaving them baffled. But the introduction of VideoPrism might change everything. It's a video encoder developed by Google's research team that can achieve state-of-the-art levels on a variety of video understanding tasks with a single model. Whether it

Breaking News! Gen-3 Alpha Text to Video is now open to everyone for use! This is a brand new field that can generate videos with high fidelity, speed, and control. Visit the official website http://runwayml.com right now to experience it! This news has sparked widespread attention, and many are eagerly waiting for the arrival of this new feature.Official VideoAccording to AIbase, the Gen-3 Alpha Text to Video service is currently available only to paid users, with an annual subscription price o

Famous Hollywood tycoon Tyler Perry has halted an $800 million investment project. He believes the newly released generative video model Sora will have a significant impact on the film industry. Sora will simplify the filmmaking process and may lead to job losses in certain positions. Tyler Perry expressed that the speed of AI technology development has shocked him, and the film industry will face challenges.

Jiang Lu, Research Lead of Google's VideoPoet project, has joined TikTok and is rumored to take on the role of head of the North American technology department. Jiang left Google to continue working in the video generation field, launching video generation technology similar to Sora. Jiang is a part-time professor at CMU and has extensive research experience, making him a key talent in AI that major companies are vying for. The VideoPoet project replaces UNet with Transformer, becoming the SOTA in AI video generation, attracting TikTok's interest.

The Sora generative video model has recently sparked significant discussion, with reports claiming that one of its inventors is Xie Saining, a young scientist who graduated from Shanghai Jiao Tong University. Xie Saining responded on social media, denying any connection to Sora and stating that he has no association with the project. He emphasized that for complex systems like Sora, talent is the most important factor, followed by data and then computational power. He also mentioned that before AI regulation and technological maturity, simply generating beautiful images is not enough; capturing the truth is essential, much like what is described in 'Black Mirror.'

Midjourney has announced plans to launch a text-to-video model in the coming months. The company will begin training the video model in January, with the CEO stating that this is a natural evolution for the platform as it introduces competitive dynamics. The latest v6 update focuses on improving image quality and user experience, highlighting the fierce competition in the AI video generation space. Entering the video domain follows the release of similar products by competitors, intensifying the competition. Midjourney is known for prioritizing quality and user experience, with the latest v6 update aimed at maintaining its competitive edge.

Midjourney announced last night that it plans to start training its video model in January. The V6 version is set to come with significant updates, including notable improvements in text processing, enhanced content coherence, and better accuracy. The focus of the V6 improvements will be the Inpainting feature, with a style adjuster potentially offering a completely new experience. The V7 version signifies a major update, indicating that Midjourney will become a strong competitor in the video generation space. The article reports on the latest developments from Midjourney, including the upcoming V6 version.

Runway has partnered with Getty Images to launch the new video model RGM, catering to businesses' needs for high-quality, customized content. The new model (RGM) provides benchmark models for businesses to build customized models for generating video content, enhancing the creative capabilities of companies across various fields. Runway and Getty Images are driving innovation in AI and video generation, enabling businesses to create professional, engaging, and brand-compliant content more easily. RGM will be open to commercial users in the coming months.