Ant Group’s Ant Technology Research Institute had four papers from its Visual Intelligence Lab accepted at NeurIPS 2022, covering rank diminishing in deep networks, geometry‑aware 3D image synthesis, dynamic discriminators for GANs, and uncertainty‑aware hierarchical refinement for incremental classification, highlighting the institute’s cutting‑edge AI research.
JD Explore’s Vega v1 model topped the GLUE benchmark with a 91.3 average score, outperforming Microsoft, Facebook, and Stanford across multiple NLP tasks, including first‑ever human‑level results on sentiment analysis and coreference, showcasing JD’s leading position in deep‑learning research.
The author shares personal experiences as a Meituan algorithm engineer, emphasizing the critical role of labeled data, the emergence of contrastive (self‑supervised) learning across computer vision, NLP, and recommendation systems, and offers practical advice for algorithm engineers to stay competitive.
AI researchAlgorithm EngineeringContrastive Learning
At ACM MM2021 in Chengdu, Kuaishou presented two accepted papers on recommendation and outfit compatibility, won the Grand Challenge with its DAP congestion‑control system, and forged a strategic partnership with the CCF‑MM committee to deepen multimedia research collaboration across academia and industry.
By constructing virtual mirror samples that occupy identical positions across source and target domains, the authors eliminate covariate shift while preserving distribution structure, enabling superior unsupervised domain adaptation that achieves state‑of‑the‑art performance on Office and VisDA benchmarks and improves real‑world lighting and gender‑recognition tasks.
This article surveys the evolution of pre‑trained models, covering the origins of transfer and self‑supervised learning, the rise of transformer‑based PTMs such as BERT and GPT, efficient architecture designs, multimodal and multilingual extensions, theoretical analyses, and future research directions for scalable and robust AI systems.
The article presents a comprehensive overview of Ant Group's intelligent service research, summarizing recent AI‑focused papers on text classification, stance detection, data augmentation, knowledge distillation for ranking, reinforcement‑learning‑based dialogue clarification, behavior‑cloning dialogue systems, anti‑fraud outbound bots, tag‑based service recommendation, and multi‑agent service groups, while also highlighting future directions and recruitment opportunities.
At ICASSP 2021, JD Tech presented three AI research papers—introducing a Neural Kalman Filtering framework for speech enhancement, a cross‑utterance BERT‑based prosody modeling method for end‑to‑end speech synthesis, and a self‑supervised conversational query rewriting approach—each demonstrating superior performance over existing baselines on benchmark datasets.
The newsletter recaps recent tech developments, including JD's four AI papers at Interspeech 2020, Shenzhen's supercomputing boost, T‑Mobile's mid‑band 5G expansion, Apple's upcoming A14T iMac processor, MongoDB Atlas multi‑cloud support, Wikimedia's migration to GitLab, and advances in graph neural network pre‑training and deep clustering.
Microsoft appointed its first Chinese Global Technical Fellow, Huang Xuedong, as the company’s Global AI CTO, overseeing Azure’s speech, translation, vision, and language services, while highlighting his groundbreaking achievements such as achieving human‑level word error rates and leading AI research teams.
This article surveys the evolution of object detection models—comparing one‑stage and two‑stage approaches, their performance trade‑offs, and recent state‑of‑the‑art methods—while also outlining key concepts and breakthroughs in natural language processing, highlighting the impact of deep‑learning models such as BERT.
A data scientist from Besedo has open‑sourced a massive NSFW image dataset containing 1.589 million pictures, organized into 159 primary categories and further sub‑categories, with download scripts and GitHub links, requiring about 500 GB of storage and cautioning against viewing in the office.
The article examines why natural language understanding is intrinsically difficult, outlines four core linguistic challenges, proposes a neural‑symbolic integration framework with three design principles, introduces the Object‑Oriented Neural Programming (OONP) architecture, and showcases real‑world applications in public security, legal document analysis, and financial fraud detection.
AI researchknowledge graphneural-symbolic integration
At the 2018 Hangzhou Yunqi Conference, Alibaba's CTO Zhang Jianfeng outlined DAMO Academy’s expansive AI strategy, unveiling new research labs, a semiconductor venture, quantum computing initiatives, and a youth science award aimed at accelerating data, algorithm, and computing breakthroughs.
The Meitu Short Video Real-Time Classification Challenge (MTSVRC), co‑hosted by the PRCV conference, Meitu and the Chinese Academy of Sciences, releases the industry‑largest dataset of over 100,000 five‑to‑15‑second videos in 50 categories, judging both classification accuracy and real‑time speed, and offers cash prizes up to ¥100,000 plus presentation opportunities for top teams.
Google Research unveiled “Talk to Books,” a semantic‑experience that lets users type statements or questions and receive full‑sentence matches from books, showcasing how AI trained on billions of conversational pairs can understand meaning beyond simple keyword matching.
The cw2vec paper, presented at AAAI 2018, introduces a Chinese word embedding method that leverages stroke n‑grams to capture character semantics, proposes a novel loss function, demonstrates consistent improvements over existing models across similarity, analogy, classification and NER tasks, and discusses real‑world AI applications.
This article reviews the key concepts, challenges, and recent research on transfer learning presented at CIKM 2017, covering instance, feature, parameter, and relation‑based methods, supervised and unsupervised deep TL approaches, and transitive transfer learning with associated loss formulations and optimization strategies.
In an exclusive interview, Alibaba AI Lab's distinguished scientist Dr. Wang Gang discusses the lab's research on multimodal AI, edge computing, AI hardware, bio‑inspired cognition, quantum‑deep‑learning integration, and the challenges of moving from recognition to true understanding, while also outlining Alibaba's AI talent recruitment plans.
Dr. Lv Zhengdong discusses the challenges of true language understanding, the integration of symbolic reasoning with neural networks, recent advances in neural‑symbolic models, and the practical prospects of NLP in domains such as law and finance, emphasizing the need for hybrid approaches.
This article explains why Alibaba's Cognitive Computing Lab uses StarCraft as a research platform, outlines the game's unique challenges for AI, and details their deep reinforcement learning and multi‑agent approaches, including the BiCNet architecture and experimental results.
StarCraft, a classic real‑time strategy game, has become a key testbed for deep reinforcement learning and AI research, and Alibaba's open‑source Gym StarCraft platform now offers Python‑based, TensorFlow‑compatible tools that simplify agent development and evaluation within the OpenAI Gym ecosystem.
The article reviews a decade of deep‑learning breakthroughs, highlights Huawei Noah's recent research on speech, image and natural‑language processing, and discusses future trends such as neural‑symbolic integration, end‑to‑end learning, and knowledge‑driven AI systems.
The article chronicles Facebook's evolution from a social platform into a leading AI research hub, detailing the founding of FAIR, its key personnel, ambitious goals, major projects such as memory networks, embedding world, DeepFace, language technology, and the M assistant, and highlights the open, collaborative nature of its AI work.