LVIS

Progress on object detection is enabled by datasets that focus the research community's attention on open challenges. This process led us from simple images to complex scenes and from bounding boxes to segmentation masks. In this work, we introduce LVIS (pronounced `el-vis'): a new dataset for Large Vocabulary Instance Segmentation. We plan to collect ~2 million high-quality instance segmentation masks for over 1000 entry-level object categories in 164k images. Due to the Zipfian distribution of categories in natural images, LVIS naturally has a long tail of categories with few training samples. Given that state-of-the-art deep learning methods for object detection perform poorly in the low-sample regime, we believe that our dataset poses an important and exciting new scientific challenge.

Papers using LVIS (24)

Grounding DINO: Marrying DINO With Grounded Pre-training For Open-set Object Detection2023 · 1,356 citesYolo-world: Real-time Open-vocabulary Object Detection2024 · 553 citesEqualization Loss For Long-tailed Object Recognition2020 · 482 citesFrustratingly Simple Few-shot Object Detection2020 · 300 citesOpen-vocabulary Object Detection Via Vision And Language Knowledge Distillation2021 · 280 citesMosaicos: A Simple And Effective Use Of Object-centric Images For Long-tailed Object Detection2021 · 39 citesX-DETR: A Versatile Architecture For Instance-wise Vision-language Tasks2022 · 37 citesScaling Open-vocabulary Object Detection2023 · 28 citesDiffusionInst: Diffusion Model for Instance Segmentation2022 · 23 citesClassification Calibration For Long-tail Instance Segmentation2019 · 9 citesBridging Images And Videos: A Simple Learning Framework For Large Vocabulary Video Object Detection2022 · 7 citesDetect Everything With Few Examples2023 · 5 citesRegion-centric Image-language Pretraining For Open-vocabulary Detection2023 · 4 citesLanguage-conditioned Detection Transformer2023 · 3 citesReal-time Transformer-based Open-vocabulary Detection With Efficient Fusion Head2024 · 3 citesSOS: Segment Object System For Open-world Instance Segmentation With Object Priors2024 · 2 citesEnhancing Novel Object Detection Via Cooperative Foundational Models2023 · 1 citesEqualization Loss For Large Vocabulary Instance Segmentation2019 · 1 citesUnsupervised Discovery of the Long-Tail in Instance Segmentation Using Hierarchical Self-Supervision2021Enhancing Open-Vocabulary Object Detection through Multi-Level Fine-Grained Visual-Language Alignment2026Moondream Segmentation: From Words to Masks2026Search And Detect: Training-free Long Tail Object Detection Via Web-image Retrieval2024Gen2det: Generate To Detect2023Logit Normalization For Long-tail Object Detection2022