Author

Bilibili Tech

Provides introductions and tutorials on Bilibili-related technologies.

403

Articles

Likes

3.0k

Views

Comments

Latest from Bilibili Tech

100 recent articles max

Bilibili Tech

Aug 8, 2025 · Artificial Intelligence

Can Language‑Centric Tree Reasoning Transform Video Question Answering?

This article introduces a language‑centric tree reasoning (LTR) framework that recursively decomposes VideoQA queries into perceptual sub‑questions and performs bottom‑up logical inference with video assistance, achieving significantly higher accuracy and explainability across eleven benchmark datasets.

Artificial IntelligenceMultimodal LLMTree Reasoning

0 likes · 17 min read

Can Language‑Centric Tree Reasoning Transform Video Question Answering?

Bilibili Tech

Aug 5, 2025 · Artificial Intelligence

How Bilibili’s IndexTTS2 Achieves Real‑Time, Emotion‑Rich Voice Translation

IndexTTS2 introduces a cross‑modal, multi‑language voice translation system that preserves speaker identity, acoustic space, and multi‑source timbre, while tackling challenges like voice personality loss, subtitle cognitive load, localization costs, multi‑speaker diarization, and cultural adaptation through novel time‑coding, adversarial RL, and diffusion‑based lip‑sync techniques.

Speech Synthesisadversarial reinforcement learningcross-language TTS

0 likes · 20 min read

How Bilibili’s IndexTTS2 Achieves Real‑Time, Emotion‑Rich Voice Translation

Bilibili Tech

Aug 1, 2025 · Mobile Development

Revamping a Mobile Video Editor: MVVM + UDF Architecture and Redo/Undo Design

This article details the comprehensive redesign of a mobile video‑editing page, covering background challenges, requirement analysis, MVVM + UDF architectural choices, module decomposition, dependency‑injection implementation, and the design of a robust Redo/Undo system to improve maintainability and user experience.

ArchitectureDependency InjectionMVVM

0 likes · 19 min read

Revamping a Mobile Video Editor: MVVM + UDF Architecture and Redo/Undo Design

Bilibili Tech

Jul 25, 2025 · Big Data

How Unified Metadata Lineage Transforms Big Data Governance and Security

This article introduces the comprehensive design and evolution of a unified metadata lineage platform for big data, covering background, data processing chain, lineage models, system architecture, quality metrics, application scenarios, and future plans to enhance data governance, quality, and security.

ArchitectureBig DataData Governance

0 likes · 27 min read

How Unified Metadata Lineage Transforms Big Data Governance and Security

Bilibili Tech

Jul 18, 2025 · Frontend Development

How We Built a Unified Frontend Material Platform to Replace cnpm and Boost Efficiency

This article describes the design and implementation of a new frontend material platform that consolidates npm packages, UMD scripts, and icons, introduces a custom publishing tool (bpm), integrates with Nexus3, adds unified authentication, improves search with MeiliSearch, and outlines migration, classification, and documentation strategies.

documentationfrontendmaterial library

0 likes · 27 min read

How We Built a Unified Frontend Material Platform to Replace cnpm and Boost Efficiency

Bilibili Tech

Jul 11, 2025 · Artificial Intelligence

IndexTTS2: Emotionally Expressive, Duration-Controlled Zero-Shot TTS

IndexTTS2 introduces a novel auto-regressive zero-shot text-to-speech model that achieves precise duration control and fine-grained emotional expression through a universal time‑encoding mechanism, decoupled voice‑style and emotion modeling, and a GPT‑style latent feature, outperforming state‑of‑the‑art baselines across multiple benchmarks.

Text-to-Speechduration controlemotional synthesis

0 likes · 23 min read

IndexTTS2: Emotionally Expressive, Duration-Controlled Zero-Shot TTS

Bilibili Tech

Jul 4, 2025 · Operations

Solving CPU Performance Layering in Heterogeneous Data Centers: A Practical Guide

This article explains why heterogeneous servers cause CPU performance layering, describes how to detect the issue using metrics such as NUMA hit/miss rates, cache miss ratios and frequency states, and provides step‑by‑step remediation techniques—including NUMA binding, cache isolation, recompilation and frequency locking—to improve resource pooling efficiency in modern data centers.

CPU performanceNUMAcache optimization

0 likes · 24 min read

Solving CPU Performance Layering in Heterogeneous Data Centers: A Practical Guide

Bilibili Tech

Jun 13, 2025 · Mobile Development

How Bilibili Scaled Kotlin Multiplatform Across Android, iOS, and HarmonyOS

This article details Bilibili's practical experience with Kotlin Multiplatform (KMP), covering the choice of Bazel as a build system, multi‑language interop, dependency injection, modular export, state‑machine driven single‑direction data flow, and the successful deployment of shared logic and UI across Android, iOS, and HarmonyOS platforms.

BazelCompose UICross‑platform

0 likes · 20 min read

How Bilibili Scaled Kotlin Multiplatform Across Android, iOS, and HarmonyOS

Bilibili Tech

May 27, 2025 · Operations

Automated Server Fault Detection and Repair: Architecture, Methods, and Future Outlook

This article presents a comprehensive overview of server fault management at scale, detailing the classification of failures, shortcomings of traditional manual processes, and the design of an automated detection and repair system that combines in‑band and out‑of‑band data collection, rule‑based alerting, and end‑to‑end repair workflows, while also outlining future directions for intelligent monitoring and reliability.

AutomationInfrastructureOperations

0 likes · 17 min read

Automated Server Fault Detection and Repair: Architecture, Methods, and Future Outlook

Bilibili Tech

May 20, 2025 · Artificial Intelligence

How AnimeReward and GAPO Transform Anime Video Generation with Human Feedback

Researchers at Bilibili present Index‑Anisora, an open‑source anime video generation framework that builds a 30k‑sample reward dataset, introduces the multi‑dimensional AnimeReward model and a Gap‑Aware Preference Optimization (GAPO) method, and demonstrate through extensive automatic and human evaluations that their approach significantly outperforms baseline video generators.

AIGAPOHuman Feedback

0 likes · 20 min read

How AnimeReward and GAPO Transform Anime Video Generation with Human Feedback