Multiple instance and context dependent learning in hyperspectral data

Torrione, Peter A.; Ratto, Christopher R.; Collins, Leslie M.

doi:10.1109/whispers.2009.5289021

Cited by 17 publications

(10 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The DP based context approaches are compared to the supervised generative context approach presented in [3], a noncontext discriminative approach and the RX prescreener. As in [3] a supervised context of each alarm was determined according to the time of day that the image was taken, these times were then divided into three discrete classes morning, afternoon and evening.…”

Section: Resultsmentioning

confidence: 99%

“…These approaches are distinguished by the manner in which context is defined and inferred. In [3] data surrounding each location of interest is used to characterize the underlying context using a generative statistical model and a different classifier is trained for each context. In this work this methodology is referred to as a generative context approach.…”

Section: Introductionmentioning

confidence: 99%

“…The latent context variables are inferred to maximize discrimination performance and therefore these approaches are known as discriminative context approaches. The generative approach in [3] is also different from the discriminative approaches in that it also requires the context of each observation be labeled a priori. It is possible to envision a similar approach in which contexts are not labeled a priori but are instead inferred in an unsupervised fashion using a probabilistic mixture model.…”

Section: Introductionmentioning

confidence: 99%

“…A DP based generative approach to context dependent learning can be realized by estimating the parameters of a DP Gaussian mixture model from the context data. The number of utilized components defines the number of contexts and context dependent classifier development can proceed as in [3]. A DP based discriminative approach to context dependent learning can be realized using a DP mixture of logistic discriminants [9].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Dirichlet process based context learning for mine detection in hyperspectral imagery

Morton

Torrione

Collins

2010

2010 2nd Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing

View full text Add to dashboard Cite

Hyperspectral imagery (HSI) has been shown to be a powerful remote sensing phenomenology that is appropriate for a variety of classification and detection tasks. Standard detection and classification algorithms applied to hyperspectral data are hindered by environmental factors that alter the statistics of the data such as sun intensity, atmospheric conditions or soil properties. Detection and classification algorithms operating on HSI must account for the changing context underlying each observation for robust performance. This work focuses on algorithms that incorporate knowledge of underlying context for the discrimination of landmine responses from other surface or sub-surface anomalies using airborne HSI. This work compares both generative context models, that model context at a given location using features of the surrounding data, and discriminative context models that determine the context at a given location to maximize performance. Both approaches utilize a Dirichlet process prior to infer the number of contexts within the data without the need to explicitly label the context of each image or location within the image. Results indicate that Dirichlet process based generative context clustering determines contexts that are congruent with physical characteristics such as time of day, but does not necessarily lead to performance improvements. Dirichlet process based discriminative clustering, however, yields performance greater than a labeled generative approach.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Dirichlet process based context learning for mine detection in hyperspectral imagery

Morton

Torrione

Collins

2010

2010 2nd Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Recently, MIL algorithms have also been applied to remote sensing image classification. For example, in [14] MIL approach is explored for sub-surface landmine detection using hyperspectral (HS) imagery. In [2], authors have developed MIL based binary classification scheme for identifying targets (landmines) in HS imagery.…”

Section: Related Work and Our Contributionsmentioning

confidence: 99%

Complex settlement pattern extraction with multi-instance learning

Vatsavai

Bhaduri

Graesser

2013

Joint Urban Remote Sensing Event 2013

View full text Add to dashboard Cite

Per-pixel (or single instance) based classification schemes which have proven to be very useful in thematic classification have shown to be inadequate when it comes to analyzing very high resolution remote sensing imagery. The main problem being that the pixel size (less than a meter) is too small as compared to the typical object size (100s of meters) and contains too little contextual information to accurately distinguish complex settlement types. One way to alleviate this problem is to consider a bigger window or patch/segment consisting a group of adjacent pixels which offers better spatial context than a single pixel. Unfortunately, this makes per-pixel based classification schemes ineffective. In this work, we look at a new class of machine learning approaches, called multi-instance learning, where instead of assigning class labels to individual instances (pixels), a label is assigned to the bag (all pixels in a window or segment). We applied this multi-instance learning approach for identifying two important urban patterns, namely formal and informal settlements. Experimental evaluation shows the better performance of multi-instance learning over several well-known single-instance classification schemes.

show abstract

A Scalable Complex Pattern Mining Framework for Global Settlement Mapping

Vatsavai

2015

2015 IEEE International Congress on Big Data

View full text Add to dashboard Cite

Multiple instance and context dependent learning in hyperspectral data

Cited by 17 publications

References 5 publications

Dirichlet process based context learning for mine detection in hyperspectral imagery

Dirichlet process based context learning for mine detection in hyperspectral imagery

Complex settlement pattern extraction with multi-instance learning

A Scalable Complex Pattern Mining Framework for Global Settlement Mapping

Contact Info

Product

Resources

About