BestHub
Sign in
Tagged articles
7 articles
Page 1 of 1
Kuaishou Large Model
Kuaishou Large Model
Sep 27, 2023 · Artificial Intelligence

DVIS: Decoupled Framework that Sets New SOTA in Video Instance Segmentation

DVIS introduces a decoupled video instance segmentation framework that splits the task into segmentation, tracking, and refinement modules, achieving state-of-the-art performance across VIS, VPS, and VSS benchmarks while maintaining low computational overhead, and demonstrates robustness in both online and offline settings.

Computer VisionDeep LearningTransformer
0 likes · 12 min read
DVIS: Decoupled Framework that Sets New SOTA in Video Instance Segmentation
Alimama Tech
Alimama Tech
Feb 1, 2023 · Artificial Intelligence

Video Object of Interest Segmentation (VOIS): Task, Dataset, and Dual-Path Transformer Approach

The paper presents Video Object of Interest Segmentation (VOIS), a new e‑commerce task that locates and segments video instances matching a given product image, introduces the LiveVideos dataset of 2,418 Taobao live‑stream clips, and proposes a dual‑path Swin‑Transformer with cross‑fusion modules that outperforms existing VOS/VIS baselines.

DatasetTransformerinstance segmentation
0 likes · 11 min read
Video Object of Interest Segmentation (VOIS): Task, Dataset, and Dual-Path Transformer Approach
AntTech
AntTech
Oct 19, 2021 · Artificial Intelligence

Target Re‑identification and Occluded Video Instance Segmentation: Applications in Insurance Claims and Pet Identification

The article introduces pet identity verification using target re‑identification and occluded video instance segmentation, describes recent ICCV VIPriors competitions where Ant Group’s insurance team achieved top ranks, and explains how these computer‑vision techniques are applied to insurance claims, pet identification, and future AI scenarios.

Insurance AITarget Re-identificationinstance segmentation
0 likes · 7 min read
Target Re‑identification and Occluded Video Instance Segmentation: Applications in Insurance Claims and Pet Identification
Kuaishou Tech
Kuaishou Tech
May 24, 2021 · Artificial Intelligence

BCNet: A Bilayer Instance Segmentation Network for Occlusion‑Aware Object Detection

The paper proposes BCNet, a lightweight bilayer instance segmentation network that explicitly models occluder and occludee relationships by treating each region of interest as two overlapping layers, achieving significant performance gains on COCO, COCOA and KINS datasets under heavy occlusion.

Computer VisionDeep Learningbilayer network
0 likes · 10 min read
BCNet: A Bilayer Instance Segmentation Network for Occlusion‑Aware Object Detection
Kuaishou Audio & Video Technology
Kuaishou Audio & Video Technology
May 21, 2021 · Artificial Intelligence

How BCNet Tackles Occlusion in Instance Segmentation with a Dual‑Layer GCN

The article introduces BCNet, a lightweight dual‑layer instance segmentation network that models images as overlapping occluder and occludee layers, enabling effective handling of heavy object occlusion and achieving significant performance gains on COCO, COCOA and KINS datasets compared to existing methods.

graph convolutional networkinstance segmentationocclusion handling
0 likes · 11 min read
How BCNet Tackles Occlusion in Instance Segmentation with a Dual‑Layer GCN
HomeTech
HomeTech
Apr 21, 2021 · Artificial Intelligence

AI-Powered Masked Danmaku: Design and Implementation

This article details the design and practical implementation of an AI-driven masked danmaku system that prevents comment overlay on video content, covering background, technology selection, instance segmentation methods, distributed task scheduling, mask generation, client rendering, performance optimizations, and future directions.

Distributed SystemsMask DanmakuVideo processing
0 likes · 18 min read
AI-Powered Masked Danmaku: Design and Implementation
Meituan Technology Team
Meituan Technology Team
Dec 24, 2020 · Artificial Intelligence

Meituan Unmanned Delivery Technical Salon – AI Research on Instance Segmentation, Visual Localization, Trajectory Prediction, and Depth‑Pose Learning

On January 9, 2021, Meituan hosted an unmanned‑delivery technical salon in Beijing where experts presented cutting‑edge AI research—including the CenterMask instance‑segmentation method, 3D geometry‑aware camera localization, multi‑agent trajectory prediction with attention‑based spatio‑temporal graphs, real‑time stereo visual‑inertial odometry calibration, and self‑supervised depth‑pose learning for dynamic scenes.

Computer Visionaiautonomous driving
0 likes · 7 min read
Meituan Unmanned Delivery Technical Salon – AI Research on Instance Segmentation, Visual Localization, Trajectory Prediction, and Depth‑Pose Learning