Multiple Instance Boosting for Object Detection

By Paul A. Viola, John C. Platt, Cha Zhang. Neural Information Processing Systems, 2005

801 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/521768f7772163bc7c57ae2c9855889abb747fbb

How researchers see this paper

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

Citations in the corpus

Two-person interaction detection using body-pose features and multiple instance learning (2012, background): “…datasets are focused on simple activities or gestures [19, 16, 2], or daily activities [22, 32] performed by a single actor such as drinking water, cooking, entering the room, etc. Acitivity recognition with Kinect: We briefly mention approaches to the single or daily activity…”
Image Captioning and Visual Question Answering Based on Attributes and External Knowledge (2016, methodology): “[37] solved a similar problem using a Multiple Instance Learning framework [61] to detect visual words from images.”
Scale-Aware Face Detection (2017, background): “Many works have been proposed to improve the Viola-Jones framework and achieves further improvements, such as local features [41, 20, 36], boosting algorithms [40, 21, 11], cascade structure [2] and multi-pose [22, 17, 12].”
PROST: Parallel robust online simple tracking (2010, background): “[22] showed that multiple-instance learning can easily handle such ambiguities of bounding boxes.”
Evaluate the Malignancy of Pulmonary Nodules Using the 3-D Deep Leaky Noisy-OR Network (2017, methodology): “[34] used it as a boosting method to improve the object detection accuracy.”
Robust Object Tracking with Online Multiple Instance Learning (2011, unknown): “This technique has found recent success in other computer vision areas, such as object detection [14], [15] and object recognition [16], [17], [18].”
SMILE: A Similarity-Based Approach for Multiple Instance Learning (2010, methodology): “In MILBoost [13], all instances initially get the same label as the bag label for training the first classifier and each instance is assigned a weight indicating its label.”
Confidence-Rated Multiple Instance Boosting for Object Detection (2014, methodology): “Applications to object detection have thus far been limited to coping with slight translational and scale uncertainties around a manually annotated bounding box [17, 4, 8].”
Visual to Text: Survey of Image and Video Captioning (2019, methodology): “Different from seeking a co-embedding representation, Pan et al. [109] detected visual attributes from frames/videos with a multiple instance learning (MIL) [114] model, and fed these attributes as high-level semantics to a two-layer stacked LSTM decoder, where the first LSTM is initialized with vid”
Visual tracking with online Multiple Instance Learning (2009, methodology): “The algorithm we propose is based on boosting and is related to the MILBoost algorithm [23] as well as the Online-AdaBoost algorithm [20] (to our knowledge no other online MIL algorithm currently exists in the literature).”
Multiple instance learning: A survey of problem characteristics and applications (2016, methodology): “MIL-Boost [54]: The MIL-Boost algorithm used in this paper is a generalization of the algorithm presented in [142].”
Patch-based Convolutional Neural Network for Whole Slide Tissue Image Classification (2015, background): “The probability of a bag being positive equals to the maximum positive prediction over all of its instances [6, 54, 27].”
Data-efficient and weakly supervised computational pathology on whole-slide images (2020, methodology): “…using max pooling or other aggregation functions ( e.g. mean operator, generalized mean, log-sum-exp, the quantile function, Noisy-OR, Nosiy-And 35–37 , etc. ) of limited ﬂexibility for dataset or problem-speciﬁc tuning, we build CLAM around the trainable, attention-based pooling function 33 to…”
Attention-based Deep Multiple Instance Learning (2018, unknown): “, 2006), and combining responses of an instance-level classifier (Ramon & De Raedt, 2000; Raykar et al., 2008; Zhang et al., 2006).”
Exploring Person Context and Local Scene Context for Object Detection (2015, background): “ored with a classiﬁer that assigns a probability p irfor each category iindicating its conﬁdence that an object of this category is present in the image. In the Noisy-Or model for image classiﬁcation [31], these probabilities are combined to give a probability to the entire image: p i= 1 Y r (1 p ir”