Tencent's official tech account. Delivering quality technical content to serve developers.
Tencent AI Lab and Shenzhen Mindray have signed an AI cooperation framework to develop an automated blood cell analysis system that leverages machine learning, computer vision, and cloud solutions, aiming to improve early disease screening, reduce costs, and address the shortage of skilled microscopy personnel.
Zhang Zhengyou, head of Tencent AI Lab and Robotics X, has become the company’s inaugural T17‑level researcher, a distinction earned through pioneering work in AI‑driven agriculture, virtual humans, facial recognition, speech front‑ends, advanced robotics like the Jamoca quadruped, and contributions such as the Go AI “Absolute Art”.
Tencent AI Lab’s WeKick AI clinched the inaugural Google Football Kaggle championship with a score of 1785.8, outpacing over 1,100 research teams by leveraging a migrated full‑stack architecture, custom reinforcement‑learning frameworks, GAIL imitation learning, and a multi‑style League training pipeline that together showcase the power and generality of deep reinforcement learning for complex multi‑agent tasks.
Tencent Cloud released a comprehensive 30,000‑word cloud‑native roadmap that explains core concepts, best‑practice stages, and real‑world case studies across industries, helping enterprises and developers assess their current cloud‑native maturity and plan effective migration paths.
This article chronicles Tencent's journey from the early development of the TFS distributed storage platform to large‑scale data migrations, flexible bandwidth strategies, and the creation of the cloud‑native YottaStore, illustrating how a small architecture team solved massive storage challenges for billions of users.
Tencent’s AI Security Attack Matrix, the industry’s first AI‑focused risk framework, maps attack tactics, techniques, and processes across the AI lifecycle, offering practical guidance for researchers and developers to identify and mitigate security threats in AI systems.
Tencent engineers broke the world record for training ImageNet with 128 V100 GPUs in just 2 minutes 31 seconds, detailing a suite of optimizations—including a new Light distributed training framework, single‑machine speed boosts, multi‑machine communication enhancements, and advanced batch convergence techniques—that together dramatically cut training time while maintaining high accuracy.
Tencent joins the Linux Foundation Public Health initiative, open‑sourcing tools like the international COVID‑19 health module, an AI self‑triage assistant, and a severe‑case prediction model with Zhong Nanshan’s team, empowering developers, hospitals, and public‑health agencies worldwide to combat the pandemic.
This article explains why VoIP calls suffer from brief interruptions caused by packet loss, compares traditional forward error correction (FEC) and packet loss concealment (PLC) techniques, introduces Tencent's deep‑learning‑enhanced cPLC, and shows how it significantly improves MOS scores, especially under burst‑loss conditions.
This article examines the rapid rise of live‑ecommerce during the 618 shopping festival, explains why the “live + ecommerce” model demands robust backend, streaming and CDN infrastructure, and details Tencent Cloud’s architectural solutions—including media processing, low‑latency protocols, bandwidth optimization and anti‑attack measures—to support massive concurrent traffic.
