Search AI Products and News

Meta has officially launched a new short video editing function, which is based on its advanced Movie Gen AI technology, providing users with a more convenient and creative video editing experience. Users can now easily replace costumes, scenes, and styles in videos, creating amazing effects that seem to open a door to a world of wonder. This new tool supports over 50 preset effects, allowing users to choose from options that can transform video characters into comic art, marble statues, and even...

Recently, AIbase obtained the latest information from social media platforms and learned about Shisa.AI, a HuggingFace model provider specializing in Japanese fine-tuning. Its newly released bilingual Japanese-English model has attracted significant attention in the industry. This article will provide you with an in-depth analysis of Shisa.AI's latest achievements and its breakthrough advancements in the field of Japanese AI. Shisa V2405B: The birth of Japan's strongest open source model According to AIbase's understanding, Shisa.AI has recently released a Llama3.1-based S

Kimi, the intelligent assistant under Moon's Dark Side Technology Co., Ltd., announced a major upgrade to its product interface. The core goal of this upgrade is to break the limitations of traditional dialog boxes, transforming Kimi from a simple chat tool into a more powerful and information-rich multimodal container.

On May 8th, OpenAI o4-mini launched with Reinforcement Fine-Tuning. The combination of these two technologies completely transforms the cost structure and technical threshold for AI specialization, enabling enterprises to rapidly turn general-purpose AI into expert systems in specific fields using only a small amount of training data. The leap from general intelligence to expert-level AI is highlighted by the Reinforcement Fine-Tuning technology at the core of this release. Unlike traditional supervised fine-tuning, RFT is based on reinforcement learning algorithms, driven by reward-based training.

JSON Visuals for ChatGPT is officially released, injecting a new creative dimension into ChatGPT's image generation capabilities. According to AIbase, this tool offers over 50 unique aesthetic codes, combined with an attribute randomizer, to generate an infinite number of style combinations. Users simply input an image and JSON style code to create personalized visual content. The release announcement has sparked enthusiastic responses on social media, with the community particularly praising its surreal tech style. Core features: Flexible style generation and randomized JSON combinations.

Ghiblio.art is an online AI tool specializing in transforming photos into artwork in the style of Studio Ghibli. Studio Ghibli is renowned for its hand-drawn style, soft color palettes, and fantastical scenes, as seen in classics like Spirited Away and My Neighbor Totoro. Using advanced AI algorithms, Ghiblio.art automatically recreates user-uploaded photos as Ghibli-esque illustrations, imbuing ordinary images with a dreamy, animated quality. AIbase learned that users simply upload a photo and select...

This project details the creation of Pusa, a high-quality open-source video model built with a budget under $100. It leverages Mochi fine-tuning for efficient and cost-effective training, demonstrating the feasibility of achieving impressive results with minimal resources.

Google Docs has rolled out a highly anticipated new feature: Audio Overviews. This feature aims to improve writing quality by allowing users to listen to their documents read aloud, offering a convenient multitasking option for busy individuals. Audio Overview reads the document's content aloud, helping authors identify spelling errors and awkward phrasing. Writers often overlook minor mistakes or stylistic issues; this feature helps address that blind spot.

DeepSeek AI, a prominent Chinese artificial intelligence research lab, following its powerful open-source language model DeepSeek-R1, has achieved another significant breakthrough in the field of Large Language Models (LLMs). Recently, DeepSeek AI officially launched an innovative technology called Self-Principled Critique Tuning (SPCT), aimed at building more general-purpose and scalable AI reward models.

A groundbreaking achievement in quantum computing has been reached with the successful fine-tuning of a billion-parameter AI model on Origin Quantum's Wukong, marking a significant milestone in the development of quantum-enhanced AI.

In the field of AI art generation, diffusion models are transitioning from U-Net based architectures to Transformer-based architectures (DiT). However, the DiT ecosystem faces challenges in plugin support, efficiency, and multi-conditional control. Recently, a team led by Xiaojiu-z introduced EasyControl, an innovative framework designed to provide efficient and flexible conditional control capabilities for DiT models, effectively giving DiT models the power of ControlNet.

Researchers from DeepSeek and Tsinghua University recently published a new paper exploring scaling methods for reward model inference, seemingly advancing DeepSeek R2. Reinforcement learning is widely used in the large-scale post-training phase of large language models, but faces the challenge of obtaining accurate reward signals for these models. The researchers found that using pointwise generative reward modeling (GRM) improves model adaptability and scalability during inference. To this end, they propose Self-Principle Calibration Tuning (SPCT) learning.

Recently, an AI model named EasyControl_Ghibli quietly launched on the Hugging Face platform, bringing exciting news to users: generate images in the style of Studio Ghibli effortlessly and without cost. This model's release provides a user-friendly tool for anime enthusiasts and creative individuals, allowing more people to incorporate their imaginations into the classic Ghibli aesthetic and brighten their daily lives. EasyControl...

In today's rapidly evolving technological landscape, Artificial Intelligence (AI) is permeating all aspects of our lives at an astonishing pace. Recently, ChatGPT introduced a highly anticipated new feature – AI image generation. This feature not only caters to the creative needs of a wide range of users but has also garnered significant popularity for its unique "Ghibli-style" output. However, the launch of this new feature has also sparked some controversy, particularly with renowned animation director Hayao Miyazaki publicly expressing his disapproval of AI-generated art. The introduction of this image generation tool marks a significant...

OpenAI's recently launched ChatGPT image generation feature has garnered significant attention on social media, particularly its ability to generate images in the "Ghibli style." Previously, this native image generation capability was only available to paid ChatGPT Plus, Pro, and Team users globally. Now, this new ChatGPT feature has quietly been opened to all free users. The 'Create image' option displayed below the chatbox has been updated, transforming ideas and concepts into visuals.