Generating professional financial reports is a labor-intensive and intellectually demanding process that current AI systems struggle to fully automate. To address this challenge, we introduce FinSight (Financial InSight), a novel multi agent framework for producing high-quality, multimodal financial reports. The foundation of FinSight is the Code Agent with Variable Memory (CAVM) architecture, which unifies external data, designed tools, and agents into a programmable variable space, enabling flexible data collection, analysis and report generation through executable code. To ensure professional-grade visualization, we propose an Iterative Vision-Enhanced Mechanism that progressively refines raw visual outputs into polished financial charts. Furthermore, a two stage Writing Framework expands concise Chain-of-Analysis segments into coherent, citation-aware, and multimodal reports, ensuring both analytical depth and structural consistency. Experiments on various company and industry-level tasks demonstrate that FinSight significantly outperforms all baselines, including leading deep research systems in terms of factual accuracy, analytical depth, and presentation quality, demonstrating a clear path toward generating reports that approach human-expert quality.

As millions of people across China travel to the graves of their ancestors to pay their respects for the annual tomb-sweeping festival, a new way of remembering, and reviving, their beloved relatives is being born. For as little as 20 yuan ( 2.20), Chinese netizens can create a moving digital avatar of their loved one, according to some services advertised online. So this year, to mark tomb-sweeping festival on Thursday, innovative mourners are turning to artificial intelligence to commune with the departed. At the more sophisticated end of the spectrum, the Taiwanese singer Bao Xiaobai used AI to "resurrect" his 22-year-old daughter, who died in 2022. Despite having only an audio recording of her speaking three sentences of English, Bao reportedly spent more than a year experimenting with AI technology before managing to create a video of his daughter singing happy birthday to her mother, which he published in January.

Large Language Models (LLMs) have demonstrated proficiency in addressing tasks that necessitate a combination of task planning and the usage of external tools that require a blend of task planning and the utilization of external tools, such as APIs. However, real-world complex systems present three prevalent challenges concerning task planning and tool usage: (1) The real system usually has a vast array of APIs, so it is impossible to feed the descriptions of all APIs to the prompt of LLMs as the token length is limited; (2) the real system is designed for handling complex tasks, and the base LLMs can hardly plan a correct sub-task order and API-calling order for such tasks; (3) Similar semantics and functionalities among APIs in real systems create challenges for both LLMs and even humans in distinguishing between them. In response, this paper introduces a comprehensive framework aimed at enhancing the Task Planning and Tool Usage (TPTU) abilities of LLM-based agents operating within real-world systems. Our framework comprises three key components designed to address these challenges: (1) the API Retriever selects the most pertinent APIs for the user task among the extensive array available; (2) LLM Finetuner tunes a base LLM so that the finetuned LLM can be more capable for task planning and API calling; (3) the Demo Selector adaptively retrieves different demonstrations related to hard-to-distinguish APIs, which is further used for in-context learning to boost the final performance. We validate our methods using a real-world commercial system as well as an open-sourced academic dataset, and the outcomes clearly showcase the efficacy of each individual component as well as the integrated framework.

Artificial Intelligence poses both risks and rewards, but developers should be weary of technologies that could threaten "scary" outcomes, AI technologist says. China's cyberspace watchdog unveiled a draft proposal Tuesday for how to manage generative artificial intelligence services ahead of a public release, Reuters reported. The Cyberspace Administration of China (CAC) said that content from generative AI services must align with the country's core socialist values. Generative AI's are bots that aim to create new things by consulting their existing body of data. In this photo illustration, SenseTime logo is seen displayed on a smartphone and a PC screen.

Text detection in natural scenes has been a significant and active research subject in computer vision and document analysis because of its wide range of applications as evidenced by the emergence of the Robust Reading Competition. One of the algorithms which has good text detection performance in the said competition is the Character Region Awareness for Text Detection (CRAFT). Employing the ICDAR 2013 dataset, this study investigates the impact of automatic image classification and blind deconvolution as image pre-processing steps to further enhance the text detection performance of CRAFT. The proposed technique automatically classifies the scene images into two categories, blurry and non-blurry, by utilizing of a Laplacian operator with 100 as threshold. Prior to applying the CRAFT algorithm, images that are categorized as blurry are further pre-processed using blind deconvolution to reduce the blur. The results revealed that the proposed method significantly enhanced the detection performance of CRAFT, as demonstrated by its IoU h-mean of 94.47% compared to the original 91.42% h-mean of CRAFT and this even outperformed the top-ranked SenseTime, whose h-mean is 93.62%.

SenseRobot, a physical artificial intelligence-powered robot, made Chinese chess history recently when it beat two professional human rivals, during a livestreamed event to an audience of 850,000. The SenseRobot AI Xiangqi Championship, which was held in Shanghai, was the first Chinese chess competition that has featured an AI-powered robot that plays Chinese chess face to face with human grandmasters. Co-hosted by leading artificial intelligence software company SenseTime, developer of SenseRobot, and Shanghai Chess Academy, the championship had the robot play against Xie Jing, a world champion, and Gu Bowen, a national youth champion. SenseRobot beat Gu at a level-16 game, while Xie failed in his challenge against the robot in a game at level 26, the most difficult level of the game. "Unlike traditional AI Chinese chess software, I was most impressed with SenseRobot's agility and steady operation, as well as its ability to calmly play the game, just like a real player," said Xie, who is an apprentice of legendary grandmaster Hu Ronghua and currently serves as a coach and player for the Shanghai Chinese Chess Team.

Novel pre-training strategy for #AI EPFL unveils a novel pre-training strategy, MultiMAE: Multi-modal Multi-task Masked Autoencoders?

Hong Kong-based AI giant SenseTime delivered a keynote titled "AI for a better tomorrow" at this year's World AI Cannes Festival (WAICF). George Huang, President of the International Business Group at SenseTime, led the keynote and showcased how the company is applying its AI technology in the areas of mixed reality, mobility, city management, and healthcare. "There has been massive market demand for AI computing services, as the digitalisation of cities and enterprises expand, along with the accelerated development of the metaverse and autonomous driving," said Huang. "We are committed to developing AI technology that advances economies, society, and humanity to address global challenges." SenseTime is the world's most-funded AI company and its facial recognition system is used for China's mass surveillance network.

BEIJING – Dozens of Chinese firms have built software that uses artificial intelligence to sort data collected on residents, amid high demand from authorities seeking to upgrade their surveillance tools, a Reuters review of government documents shows. According to more than 50 publicly available documents examined by Reuters, dozens of entities in China have over the past four years bought such software, known as "one person, one file." The technology improves on existing software, which simply collects data but leaves it to people to organize. "The system has the ability to learn independently and can optimize the accuracy of file creation as the amount of data increases. Henan's department of public security did not respond to requests for comment about the system and its uses.

"SOUTH OF THE Huai river few geese can be seen through the rain and snow." In classical Chinese this verse is a breakthrough--not in literature but in computing power. The line, composed by an artificial intelligence (AI) language model called Wu Dao 2.0, is indistinguishable in metre and tone from ancient poetry. The lab that built the software, the Beijing Academy of Artificial Intelligence (BAAI), challenges visitors to its website to distinguish between Wu Dao and flesh-and-blood 8th-century masters. Anecdotal evidence suggests that it fools most testers.