Open-Vocabulary Object Detection Using Captions

By Alireza Zareian, Kevin Dela Rosa, Derek Hao Hu, Shih-Fu Chang. 2020

613 citations recorded by the source corpus.

Despite the remarkable accuracy of deep neural networks in object detection, they are costly to train and scale due to supervision requirements. Particularly, learning more object categories typically requires proportionally more bounding box annotations. Weakly supervised and zero-shot learning techniques have been explored to scale object detectors to more categories with less supervision, but they have not been as successful and widely adopted as supervised models. In this paper, we put forth a novel formulation of the object detection problem, namely open-vocabulary object detection, which is more general, more practical, and more effective than weakly supervised and zero-shot approaches. We propose a new method to train object detectors using bounding box annotations for a limited set of object categories, as well as image-caption pairs that cover a larger variety of objects at a significantly lower cost. We show that the proposed method can detect and localize objects for which no bounding box annotation is provided during training, at a significantly higher accuracy than zero-shot approaches. Meanwhile, objects with bounding box annotation can be detected almost as accurately as supervised methods, which is significantly better than weakly supervised baselines. Accordingly, we establish a new state of the art for scalable object detection.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Multimodal Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Jianfeng Gao
2. Mohit Bansal
3. Zhe Gan
4. Caiming Xiong
5. Quoc V. Le
6. Trevor Darrell
7. Percy Liang
8. Dacheng Tao

In Computer Vision and Pattern Recognition

1. Luc Van Gool
2. Dacheng Tao
3. Ming-Hsuan Yang
4. Trevor Darrell
5. Xiaogang Wang
6. Jitendra Malik
7. Cordelia Schmid
8. Andrew Zisserman

In Multimodal Learning

1. Jianfeng Gao
2. Lijuan Wang
3. Jianwei Yang
4. Zhe Gan
5. Zicheng Liu
6. Xiaowei Hu
7. Chunyuan Li
8. Pengchuan Zhang

Citations in the corpus

• A Simple Framework for Open-Vocabulary Segmentation and Detection (2023, background): “Similarly, some open-vocabulary detection models directly leverage foundation models for distillation or transfer like OV-DETR [54] and VILD [13].”
• Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection (2023, methodology): “Grounding [62] FasterRCNN[39] - sentence partiallabel partiallabel FindIt[21] FasterRCNN[39] A sentence partiallabel fine-tune MDETR[18] DETR[2] A,C word fine-tune zero-shot fine-tune DQ-DETR[46] DETR[2] A,C word zero-shot fine-tune GLIP[26] DyHead[5] A word zero-shot zero-shot zero-shot GLIPv2[59] DyHe”
• ELEVATER: A Benchmark and Toolkit for Evaluating Language-Augmented Visual Models (2022, background): “Openset recognition allows new concepts in evaluation, while typically remains the same visual domain [87, 93]; Domain shift allows new visual domain in evaluation, while typically remains the same concept pool [63, 28].”
• RegionCLIP: Region-based Language-Image Pretraining (2021, methodology): “Zeroshot object detection aims at detecting novel object classes that are not seen during detector training [2, 18, 38, 39, 60, 65].”
• Grounded Language-Image Pre-training (2021, background): “Our setting differs from zero-shot detection [1, 13, 43, 44, 63, 68], where some categories are defined as unseen/rare and not present in the training set.”
• Open-vocabulary Object Detection via Vision and Language Knowledge Distillation (2021, methodology): “We evaluate the transferability of ViLD on PASCAL VOC (Everingham et al., 2010), COCO (Lin et al., 2014), and Objects365 (Shao et al., 2019).”
• Simple Open-Vocabulary Object Detection with Vision Transformers (2022, methodology): “OVR-CNN [42] is most similar to our approach in that it fine-tunes an image-text model to detection on a limited vocabulary and relies on image-text pre-training for generalization to an open vocabulary.”
• Detecting Twenty-thousand Classes using Image-level Supervision (2022, methodology): “Many existing methods use additional data only for pretraining [10, 66, 67], while we use image-labeled data for co-training.”
• Learning to Prompt for Open-Vocabulary Object Detection with Vision-Language Model (2022, methodology): “The representative solutions include OVR-CNN [37] and ViLD [7].”
• Open-Vocabulary Image Segmentation (2021, methodology): “Such a scenario has been recently developed and popularized for image classification [22, 38] and object detection [17, 50].”
• OpenOOD: Benchmarking Generalized Out-of-Distribution Detection (2022, background): “Future Work In the future, apart from maintaining the codebase, it is promising to extend our benchmark towards more robust OOD detectors [82] and object-level OOD detection and generalization [69, 83, 84, 85, 86, 87], which provides finer-grained visual guidance in safety-critical applications, suc”
• X-DETR: A Versatile Architecture for Instance-wise Vision-Language Tasks (2022, methodology): “First, X-DETR is evaluated on OVOD task [14, 37, 56].”
• Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models (2023, methodology): “Following [23, 27, 89], we employ a grounding loss to supervise the prediction of the masks’ category labels.”
• F-VLM: Open-Vocabulary Object Detection upon Frozen Vision and Language Models (2022, positive): “We choose Mask R-CNN per existing works (Gu et al., 2022; Zareian et al., 2021; Zhong et al., 2022).”
• Webly Supervised Concept Expansion for General Purpose Vision Models (2022, background): “Recent works build open vocabulary detectors [23, 32, 88] by leveraging image caption pairs (or models like CLIP [63] which are built from the same), obtained in large quantities on the web.”
• Learning to Generate Scene Graph from Natural Language Supervision (2021, methodology): “We anticipate that our method will benefit from the development of open-vocabulary detectors [58, 61].”
• Open Vocabulary Object Detection with Pseudo Bounding-Box Labels (2021, methodology): “Building on top of such methods, open vocabulary object detection [34] aims to improve the detection performance of novel objects with the help of image captions.”
• GLIPv2: Unifying Localization and Vision-Language Understanding (2022, background): “A long line of research focuses on this generalization problem, under the name of zero-shot [4, 62, 7, 63], weakly-supervised [18, 5, 52], or open-vocabulary [58, 19] localization.”
• A Cookbook of Self-Supervised Learning (2023, background): “These features are also leveraged by recent works for open-vocabulary object detection [Kamath et al., 2021, Gu et al., 2021, Zareian et al., 2021, Minderer et al., 2022].”