Abstract: A computer-implemented method, system and computer program product for compressing video. A set of video frames is partitioned into two subsets of different types of frames, a first type and a second type. The first type of frames of videos is compressed to generate a first representation by a first stage encoder. The first representation is then decoded to reconstruct the first type of frames using a first stage decoder. The second type of frames of video is compressed to generate a second representation that only contains soft edge information by a second stage encoder. A generative model corresponding to a second stage decoder is then trained using the first representation and the reconstructed first type of frames by using a discriminator employed by a machine learning system. After training the generative model, it generates reconstructed first and second types of frames using the soft edge information.

Abstract: A computer-implemented method, system and computer program product for compressing video. A set of video frames is partitioned into two subsets of different types of frames, a first type and a second type. The first type of frames of videos is compressed to generate a first representation by a first stage encoder. The first representation is then decoded to reconstruct the first type of frames using a first stage decoder. The second type of frames of video is compressed to generate a second representation that only contains soft edge information by a second stage encoder. A generative model corresponding to a second stage decoder is then trained using the first representation and the reconstructed first type of frames by using a discriminator employed by a machine learning system. After training the generative model, it generates reconstructed first and second types of frames using the soft edge information.

Abstract: A computer-implemented method, system and computer program product for compressing video. A set of video frames is partitioned into two subsets of different types of frames, a first type and a second type. The first type of frames of videos is compressed to generate a first representation by a first stage encoder. The first representation is then decoded to reconstruct the first type of frames using a first stage decoder. The second type of frames of video is compressed to generate a second representation that only contains soft edge information by a second stage encoder. A generative model corresponding to a second stage decoder is then trained using the first representation and the reconstructed first type of frames by using a discriminator employed by a machine learning system. After training the generative model, it generates reconstructed first and second types of frames using the soft edge information.

Abstract: A computer-implemented method, system and computer program product for compressing video. A set of video frames is partitioned into two subsets of different types of frames, a first type and a second type. The first type of frames of videos is compressed to generate a first representation by a first stage encoder. The first representation is then decoded to reconstruct the first type of frames using a first stage decoder. The second type of frames of video is compressed to generate a second representation that only contains soft edge information by a second stage encoder. A generative model corresponding to a second stage decoder is then trained using the first representation and the reconstructed first type of frames by using a discriminator employed by a machine learning system. After training the generative model, it generates reconstructed first and second types of frames using the soft edge information.

Abstract: A method, system and computer program product for measuring non-uniformity noise produced in images or videos (e.g., infrared images or videos). Images or videos, such as infrared images or videos, are captured. A model of scene statistics (statistical model of pictures, images or videos representative of pictures, images or videos, respectively, that are captured of the physical world) is utilized to measure the non-uniformity noise in the captured images or videos by exploiting exhibited characteristics for non-uniformity noise in the captured images or videos. A number signifying a magnitude of non-uniformity for each image or video frame is then generated. In this manner, non-uniformity noise produced in images or videos is measured.

Abstract: A method, system and computer program product for classifying an image or video. An image or video to be classified is received. Scene statistics (statistical model of pictures, images or videos representative of pictures, images or videos, respectively, that are captured of the physical world) of the image or video are captured. A model (a statistical model that describes a set of probability distributions) of the image or video is then created using the captured scene statistics. A comparison between the model of the image or video with two other models of images or videos is performed, such as a model of visible light images or videos and a model of infrared images or videos. The received image or video is then classified (e.g., classified as corresponding to a visible light image) based on the comparison.

Abstract: A method, system and computer program product for assessing quality of images or videos. A quality assessment of an image or video to be processed is performed using a no-reference reference quality assessment algorithm. A quality measurement, such as a score, reflecting the quality of the image or video, is generated from the no-reference reference quality assessment algorithm. The image or video is then processed and a quality assessment of the processed image or video is performed using a reference quality assessment algorithm that is conditional on the quality measurement provided by the no-reference quality assessment algorithm. In this manner, a more accurate quality measurement of the image or video is provided by the reference quality assessment algorithm.

Abstract: A method, system and computer program product for improving the recording of classroom lectures or other such presentations. A video frame containing a whiteboard image is converted into a black and white image for the detection of boundaries. These boundaries are classified as horizontal or vertical lines. Quadrangles are then formed using spatial arrangements of these lines. The quadrangles that are most likely to spatially coincide with the boundaries of the whiteboard image are identified. The quadrangles are then sorted (ranked) based on specific characteristics, such as size and position. The area corresponding to the identified quadrangle in the video frame is then cropped. Furthermore, the speaker in the video frame can be removed based on detecting changes that are characteristic of movements of a speaker. In this manner, the visual and educational experience involved in the recording of classroom lectures or other such presentation is improved.

Abstract: A method, system and computer program product for measuring non-uniformity noise produced in images or videos (e.g., infrared images or videos). Images or videos, such as infrared images or videos, are captured. A model of scene statistics (statistical model of pictures, images or videos representative of pictures, images or videos, respectively, that are captured of the physical world) is utilized to measure the non-uniformity noise in the captured images or videos by exploiting exhibited characteristics for non-uniformity noise in the captured images or videos. A number signifying a magnitude of non-uniformity for each image or video frame is then generated. In this manner, non-uniformity noise produced in images or videos is measured.

Abstract: A method, system and computer program product for predicting a viewer's quality of experience while watching mobile videos potentially afflicted with network induced impairments. The length of a stall on a video at time t is received as a first input to a model. The number of stalls up to the time t is received as a second input to the model. Furthermore, the time since a preceding rebuffering event is received as a third input to the model. Additionally, a reciprocal stall density at time t is received as a fourth input to the model. The hysteresis effect is captured using a machine-learning-based model with an input that is an aggregate of the outputs of the first, second, third and fourth inputs to nonlinear input blocks of the model, where the hysteresis effect represents an effect that a viewer's recent level of satisfaction/dissatisfaction has on their overall viewing experience.

Abstract: A method, system and computer program product for assessing quality of images or videos. A quality assessment of an image or video to be processed is performed using a no-reference reference quality assessment algorithm. A quality measurement, such as a score, reflecting the quality of the image or video, is generated from the no-reference reference quality assessment algorithm. The image or video is then processed and a quality assessment of the processed image or video is performed using a reference quality assessment algorithm that is conditional on the quality measurement provided by the no-reference quality assessment algorithm. In this manner, a more accurate quality measurement of the image or video is provided by the reference quality assessment algorithm.

Abstract: A method, system and computer program product for improving the recording of classroom lectures or other such presentations. A video frame containing a whiteboard image is converted into a black and white image for the detection of boundaries. These boundaries are classified as horizontal or vertical lines. Quadrangles are then formed using spatial arrangements of these lines. The quadrangles that are most likely to spatially coincide with the boundaries of the whiteboard image are identified. The quadrangles are then sorted (ranked) based on specific characteristics, such as size and position. The area corresponding to the identified quadrangle in the video frame is then cropped. Furthermore, the speaker in the video frame can be removed based on detecting changes that are characteristic of movements of a speaker. In this manner, the visual and educational experience involved in the recording of classroom lectures or other such presentation is improved.

Abstract: A method, system and computer program product for classifying an image or video. An image or video to be classified is received. Scene statistics (statistical model of pictures, images or videos representative of pictures, images or videos, respectively, that are captured of the physical world) of the image or video are captured. A model (a statistical model that describes a set of probability distributions) of the image or video is then created using the captured scene statistics. A comparison between the model of the image or video with two other models of images or videos is performed, such as a model of visible light images or videos and a model of infrared images or videos. The received image or video is then classified (e.g., classified as corresponding to a visible light image) based on the comparison.

Abstract: A method, system and computer program product for predicting a viewer's quality of experience while watching mobile videos potentially afflicted with network induced impairments. The length of a stall on a video at time t is received as a first input to a model. The number of stalls up to the time t is received as a second input to the model. Furthermore, the time since a preceding rebuffering event is received as a third input to the model. Additionally, a reciprocal stall density at time t is received as a fourth input to the model. The hysteresis effect is captured using a machine-learning-based model with an input that is an aggregate of the outputs of the first, second, third and fourth inputs to nonlinear input blocks of the model, where the hysteresis effect represents an effect that a viewer's recent level of satisfaction/dissatisfaction has on their overall viewing experience.

Abstract: Techniques and structures are disclosed in which one or more distortion categories are identified for an image or video, and a quality of the image or video is determined based on the one or more distortion categories. The image or video may be of a natural scene, and may be of unknown provenance. Identifying a distortion category and/or determining a quality may be performed without any corresponding reference (e.g., undistorted) image or video. Identifying a distortion category may be performed using a distortion classifier. Quality may be determined with respect to a plurality of human opinion scores that correspond to a particular distortion category to which an image or video of unknown provenance is identified as belonging. Various statistical methods may be used in performing said identifying and said determining, including use of generalized Gaussian distribution density models and natural scene statistics.

Abstract: A method, system and computer program product for improving the perceptual quality and naturalness of an image captured by an image acquisition device (e.g., digital camera). Statistical features of a scene being imaged by the image acquisition device are derived from models of natural images. These statistical features are measured and mapped onto the control parameters (e.g., exposure, ISO) of the digital acquisition device. By mapping these statistical features onto the control parameters, the perceptual quality and naturalness of the scene being imaged may be based on the values of these control parameters. As a result, these control parameters are modified to maximize the perceptual quality and naturalness of the scene being imaged. After modification of these control parameters, the image is captured by the image acquisition device. In this manner, the perceptual quality and naturalness of the image captured by the image acquisition device is improved.

Abstract: A method, system and computer program product for improving the perceptual quality and naturalness of an image captured by an image acquisition device (e.g., digital camera). Statistical features of a scene being imaged by the image acquisition device are derived from models of natural images. These statistical features are measured and mapped onto the control parameters (e.g., exposure, ISO) of the digital acquisition device. By mapping these statistical features onto the control parameters, the perceptual quality and naturalness of the scene being imaged may be based on the values of these control parameters. As a result, these control parameters are modified to maximize the perceptual quality and naturalness of the scene being imaged. After modification of these control parameters, the image is captured by the image acquisition device. In this manner, the perceptual quality and naturalness of the image captured by the image acquisition device is improved.

Abstract: Techniques and structures are disclosed in which one or more distortion categories are identified for an image or video, and a quality of the image or video is determined based on the one or more distortion categories. The image or video may be of a natural scene, and may be of unknown provenance. Identifying a distortion category and/or determining a quality may be performed without any corresponding reference (e.g., undistorted) image or video. Identifying a distortion category may be performed using a distortion classifier. Quality may be determined with respect to a plurality of human opinion scores that correspond to a particular distortion category to which an image or video of unknown provenance is identified as belonging. Various statistical methods may be used in performing said identifying and said determining, including use of generalized Gaussian distribution density models and natural scene statistics.

Abstract: Techniques and structures are disclosed in which one or more distortion categories are identified for an image or video, and a quality of the image or video is determined based on the one or more distortion categories. The image or video may be of a natural scene, and may be of unknown provenance. Identifying a distortion category and/or determining a quality may be performed without any corresponding reference (e.g., undistorted) image or video. Identifying a distortion category may be performed using a distortion classifier. Quality may be determined with respect to a plurality of human opinion scores that correspond to a particular distortion category to which an image or video of unknown provenance is identified as belonging. Various statistical methods may be used in performing said identifying and said determining, including use of generalized Gaussian distribution density models and natural scene statistics.

Abstract: Techniques and structures are disclosed in which one or more distortion categories are identified for an image or video, and a quality of the image or video is determined based on the one or more distortion categories. The image or video may be of a natural scene, and may be of unknown provenance. Identifying a distortion category and/or determining a quality may be performed without any corresponding reference (e.g., undistorted) image or video. Identifying a distortion category may be performed using a distortion classifier. Quality may be determined with respect to a plurality of human opinion scores that correspond to a particular distortion category to which an image or video of unknown provenance is identified as belonging. Various statistical methods may be used in performing said identifying and said determining, including use of generalized Gaussian distribution density models and natural scene statistics.