Article

DiscoverNet: Unsupervised Video Object Discovery through Discrete Visual Tokenization and Language Model-Inspired Reconstruction

2024-11-05

Abstract excerpt

<title>Abstract</title> <p>Object discovery, the task of separating objects from the background without human annotations, continues to be an unsolved challenge in the field of computer vision. Existing methods face difficulties due to object-background ambiguity and slot drift, as they rely on clustering based on only low-level features. In this work, we introduce DiscoverNet, a comprehensive framework that enha...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: 84817b2a-d41f-5ed8-95b9-29b10236e73c
DOI: 10.21203/rs.3.rs-5310579/v1

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

DiscoverNet: Unsupervised Video Object Discovery through Discrete Visual Tokenization and Language Model-Inspired ReconstructionDOI 10.21203/rs.3.rs-5310579/v1Read

Tarsier2: Advancing Large Vision-Language Models from Detailed Video Descriptions to Comprehensive Video UnderstandingSemantic proximity Read

Abstracted Gaussian Prototypes for One-Shot Concept LearningSemantic proximity Read

CAT: Content-Adaptive Image TokenizationSemantic proximity Read

Contextual Synergy through Explicit and Implicit Relations: A Unified Perspective for Image Description GenerationSemantic proximity Read

HEADHUNTER: Training-Free Annotated Dataset Synthesis via Self-Guided Diffusion Transformer Attention Head SelectionSemantic proximity Read

LinGen-Uni: A Universal Linear-Complexity Framework for High-Resolution Minute-Length Text-to-Video GenerationSemantic proximity Read

Real-Time Sports Action Recognition Using a CNN–Transformer Hybrid Deep Learning FrameworkSemantic proximity Read

Continuous Video Process: Modeling Videos as Continuous Multi-Dimensional Processes for Video PredictionSemantic proximity Read

Contextual Feature Expansion with Superordinate Concept for Compositional Zero-Shot LearningSemantic proximity Read

TransMODAL: A Dual-Stream Transformer with Adaptive Co-Attention for Efficient Human Action RecognitionSemantic proximity Read

Adaptive Semantic Fusion for Contextual Image CaptioningSemantic proximity Read