#### DMCA

## An information theoretic video quality metric based on motion models (2007)

Venue: | Proc. Third International Workshop on Video Processing and Quality Metrics for Consumer Electronics |

Citations: | 10 - 3 self |

### Citations

777 |
Uncertainty relation for resolution in space, spatial frequency, and orientation optimized by two-dimensional visual cortical filters
- Daugman
- 1985
(Show Context)
Citation Context .... Gabor filters attain the theoretical lower bound on the uncertainty in the frequency and spatial variables and thus, visual neurons can be said to optimize the uncertainty in information resolution =-=[12]-=-. Additionally, development of the video quality metric in Section 4.3 requires estimation of the optical flow vectors. Gabor filters have been successfully used for this purpose in the literature [13... |

363 |
Computation of component image velocity from local phase information,
- Fleet, Jepson
- 1990
(Show Context)
Citation Context ...12]. Additionally, development of the video quality metric in Section 4.3 requires estimation of the optical flow vectors. Gabor filters have been successfully used for this purpose in the literature =-=[13]-=-. A Gabor filter f(x, y, t) is simply the product of a Gaussian window and a complex exponential: f(x, y, t) = 1 ( √ 2π) 3 σxσyσt e − x 2 2σ2 + x y2 2σ2 + y t2 2σ2 ! t i(Ux+V y+W t) e (5) where (U, V,... |

249 | Model of human visual-motion sensing. - Watson, Ahumada - 1985 |

235 | Image compression via joint statistical characterization in the wavelet domain
- Buccigrossi, Simoncelli
- 1999
(Show Context)
Citation Context ...t strong dependencies along neighboring spatial locations, scales, orientations. The Gaussian Scale Mixture (GSM) model nicely describes the wavelet coefficients distributions of natural still images =-=[10, 11]-=-, which motivated the development of our information theoretic metrics for still image quality assessment [2, 3]. A continuous-frequency version of the GSM model can be posed as follows. Given a natur... |

174 |
Optic flow using spatiotemporal filters.
- Heeger
- 1988
(Show Context)
Citation Context ...xtracted using a 7 × 7 × 7 window from Ci(x) and Di(x) centered on the pixel location x. The integral in Eq. 2 can be evaluated in closed form. This integral was computed for complex Gabor filters in =-=[14]-=-. We used sine phase Gabor filters, which have zero DC response, in our implementation and evaluated this integral.sZ ∞ Z ∞ −∞ −∞ | ˜ Fs(u, v, −λxu − λyv)| 2 dudv = K(T1 − T2) T1 = e − „ −σ 2 uσ2 vσ2 ... |

171 | Scale mixtures of gaussians and the statistics of natural images
- Wainwright, Simoncelli
(Show Context)
Citation Context ...t strong dependencies along neighboring spatial locations, scales, orientations. The Gaussian Scale Mixture (GSM) model nicely describes the wavelet coefficients distributions of natural still images =-=[10, 11]-=-, which motivated the development of our information theoretic metrics for still image quality assessment [2, 3]. A continuous-frequency version of the GSM model can be posed as follows. Given a natur... |

166 | Bovik,” Video Quality Assessment Based on Structural Distortion Measurement” Signal Processing:
- Wang, Lu, et al.
- 2004
(Show Context)
Citation Context ...that the IFC is not a normalized metric [2]. The V-IFC, as specified in Eq.(9), Prediction Model SROCC Peak Signal to Noise Ratio 0.786 Proponent P8 (Swisscom) 0.803 Structural Distortion Measurement =-=[15]-=- 0.812 Visual Information Fidelity [7] 0.849 Proposed V-IFC Metric 0.876 Table 1. Comparison of SROCC values for different video quality assessment algorithms. depends not only on the distortion param... |

117 | S.G.: An information fidelity criterion for image quality assessment using natural scene statistics. - Sheikh, Bovik, et al. - 2005 |

113 | A perceptual distortion metric for digital color video. In:
- Winkler
- 1999
(Show Context)
Citation Context ...se effects. Examples of video quality metrics based on the HVS-based philosophy include the Digital Video Quality (DVQ) metric [4], the Sarnoff JND model [5] and the Perceptual Distortion Model (PDM) =-=[6]-=-. However, studies conducted by the Video Quality Experts Group indicate that the performance of HVS-based VQA algorithms leaves considerable room for improvement [1]. HVS-based VQA metrics suffer fro... |

24 | A visual information fidelity approach to video quality assessment.
- Bovik
- 2005
(Show Context)
Citation Context ...riate cortex, which play a very important role in motion perception, is not accounted for in any of these models. Preliminary extensions of the information fidelity approach has been proposed for VQA =-=[7]-=- using a simple implementation of the still image quality metric on spatio-temporal video blocks. However, this metric does not utilize motion information or model temporal artifacts in video that can... |

21 |
McGowan III. Digital video quality metric based on human vision
- Watson, Hu, et al.
- 2001
(Show Context)
Citation Context ...t system, as well as the computational model that is used to describe these effects. Examples of video quality metrics based on the HVS-based philosophy include the Digital Video Quality (DVQ) metric =-=[4]-=-, the Sarnoff JND model [5] and the Perceptual Distortion Model (PDM) [6]. However, studies conducted by the Video Quality Experts Group indicate that the performance of HVS-based VQA algorithms leave... |

19 |
Image information and visual quality,” Image Processing,
- Sheikh, Bovik
- 2006
(Show Context)
Citation Context ...le room for improvement [1]. Although progress in the development of accurate and reliable VQA algorithms has been slow, great strides have recently been made in assessing the quality of still images =-=[2, 3]-=-. In this paper, we develop a full reference quality metric for video signals by making natural extensions of the powerful information fidelity framework for still images to the spatio-temporal (video... |

3 | Statistical video models and their applications to quality assessment
- Seshadrinathan, Bovik
- 2006
(Show Context)
Citation Context ...ally filtered using a family of band-pass filters. We use a model for the statistics of these sub-band filtered coefficients that we have developed previously to assess the quality of video sequences =-=[9]-=-. This model is summarized in the next section. 2.1. Statistical Model for Video We previously proposed a model that describes the statistics of natural video sequences, which we briefly summarize her... |