Official Baidu App Tech Account
This article explains how Baidu App leverages a VDSR‑based deep‑learning model and a series of mobile‑side optimizations to deliver real‑time image and video super‑resolution on iOS and Android devices, detailing the technical challenges, performance gains, and integration steps.
The article describes Baidu App's real-time mobile super-resolution using a VDSR-based model with pruning and depthwise separable convolutions, optimized via application-layer and inference engine techniques to halve latency and memory, enabling on-device high‑def image/video enhancement, reducing server load, and supporting iOS/Android integration.
Baidu smart mini‑programs let developers embed low‑cost VR components—such as panoramic video, image, 3‑D ring and model viewers—into apps via dynamic libraries, supporting live streaming, education and marketing use cases, while Baidu’s rendering engine and upcoming 5G bandwidth aim to broaden immersive experiences.
This article explains the background of on‑device AI, introduces the Mobile Machine Learning (MML) framework and its layered architecture, details the core utilities such as model decryption and task scheduling, and provides a step‑by‑step code guide for initializing, preprocessing, inference, post‑processing, and releasing resources on mobile platforms.
Paddle.js is Baidu’s lightweight JavaScript inference engine that transforms Paddle models into web‑compatible formats, enabling fast, privacy‑preserving AI features such as face detection, gesture recognition and content filtering directly in browsers via WebGL/WebAssembly, with only 201 KB code and broad compatibility.
This article details the evolution of Baidu’s smart mini‑program framework, explains its multi‑layer architecture, outlines the startup phases and performance metrics, and provides concrete optimization techniques—including package splitting, resource compression, early data fetching, and setData best practices—to help developers deliver near‑native mobile experiences.
This guide explains why Android projects need an automated build tool, walks through the default Gradle project layout, clarifies modern dependency configurations and conflict resolution, and details the key packaging tasks, enabling developers to diagnose sync problems, manage libraries, and build reliable APKs efficiently.
By redefining performance metrics, analyzing bottlenecks, and applying targeted front‑end code and engineering optimizations—including early iOS onload, reduced first‑screen dependencies, merged and pre‑loaded APIs, bundle splitting, and modern‑mode delivery—the Baidu App personal homepage achieved over 100 ms reductions in white‑screen and first‑screen times across iOS and Android.
The article examines Android cold‑start performance by comparing three process‑creation timestamps—Application <init>, Process.getStartElapsedRealTime, and /proc/self/stat starttime—detailing their origins, extraction methods, conversion formulas, and practical guidance on which to use for accurate startup measurement.
To solve Android’s screen‑fragmentation‑induced TextView line‑spacing inconsistencies, Baidu’s team analyzed the default line‑height calculation, then created a custom LineHeightSpan (ExcludeInnerLineSpaceSpan) applied through a subclassed ETextView to enforce visual‑defined line height, delivering a simple, low‑intrusion, cross‑resolution fix.
