S. Mohammad Mostafavi I. | Gwangju Institute of Science and Technology (original) (raw)
Papers by S. Mohammad Mostafavi I.
2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019
Event cameras have a lot of advantages over traditional cameras, such as low latency, high tempor... more Event cameras have a lot of advantages over traditional cameras, such as low latency, high temporal resolution, and high dynamic range. However, since the outputs of event cameras are the sequences of asynchronous events over time rather than actual intensity images, existing algorithms could not be directly applied. Therefore, it is demanding to generate intensity images from events for other tasks. In this paper, we unlock the potential of event camera-based conditional generative adversarial networks to create images/videos from an adjustable portion of the event data stream. The stacks of space-time coordinates of events are used as inputs and the network is trained to reproduce images based on the spatio-temporal intensity changes. The usefulness of event cameras to generate high dynamic range (HDR) images even in extreme illumination conditions and also non blurred images under rapid motion is also shown. In addition, the possibility of generating very high frame rate videos is demonstrated, theoretically up to 1 million frames per second (FPS) since the temporal resolution of event cameras are about 1 µs. Proposed methods are evaluated by comparing the results with the intensity images captured on the same pixel grid-line of events using online available real datasets and synthetic datasets produced by the event camera simulator.
An event camera detects per-pixel intensity difference and produces asynchronous event stream wit... more An event camera detects per-pixel intensity difference and produces asynchronous event stream with low latency, high dynamic range, and low power consumption. As a trade-off, the event camera has low spatial resolution. We propose an end-to-end network to reconstruct high resolution, high dynamic range (HDR) images directly from the event stream. We evaluate our algorithm on both simulated and real-world sequences and verify that it captures fine details of a scene and outperforms the combination of the state-of-the-art event to image algorithms with the state-of-the-art super resolution schemes in many quantitative measures by large margins. We further extend our method by using the active sensor pixel (APS) frames or reconstructing images iteratively.
2010 International Conference on Computer Applications and Industrial Electronics, 2010
In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters appli... more In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters applied to printed text and handwriting images. CL Filters and their associated Coordinate Logic Operations (CLOs) are widely used in common practical image process applications like noise removing, magnification, opening, closing, skeletonization, coding, edge detection, feature extraction, and fractal modeling. Using coordinate logic filters increases the efficiency, simplicity and more than all the programs speed compared to morphological filters. Unlike other methods like applying Wiener filter or Median filter that are dependent to the noise applied (Gaussian or salt & pepper noise) and to the type of the input image, (handwritten text images or printed text images) coordinate logic filters are not only independent to these variations but also on the other hand accurate and fast.
2010 International Conference on Computer Applications and Industrial Electronics, 2010
In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters appli... more In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters applied to printed text and handwriting images. CL Filters and their associated Coordinate Logic Operations (CLOs) are widely used in common practical image process applications like noise removing, magnification, opening, closing, skeletonization, coding, edge detection, feature extraction, and fractal modeling. Using coordinate logic filters increases the efficiency, simplicity and more than all the programs speed compared to morphological filters. Unlike other methods like applying Wiener filter or Median filter that are dependent to the noise applied (Gaussian or salt & pepper noise) and to the type of the input image, (handwritten text images or printed text images) coordinate logic filters are not only independent to these variations but also on the other hand accurate and fast.
An event camera detects per-pixel intensity difference and produces asynchronous event stream wit... more An event camera detects per-pixel intensity difference and produces asynchronous event stream with low latency, high dynamic range, and low power consumption. As a trade-off, the event camera has low spatial resolution. We propose an end-to-end network to reconstruct high resolution, high dynamic range (HDR) images directly from the event stream. We evaluate our algorithm on both simulated and real-world sequences and verify that it captures fine details of a scene and outperforms the combination of the state-of-the-art event to image algorithms with the state-of- the-art super resolution schemes in many quantitative measures by large margins. We further extend our method by using the active sensor pixel (APS) frames or reconstructing images iteratively.
Event cameras have a lot of advantages over traditional cameras, such as low latency, high tempor... more Event cameras have a lot of advantages over traditional cameras, such as low latency, high temporal resolution, and high dynamic range. However, since the outputs of event cameras are the sequences of asynchronous events over time rather than actual intensity images, existing algorithms could not be directly applied. Therefore, it is demanding to generate intensity images from events for other tasks. In this paper, we unlock the potential of event camera-based conditional generative adversarial networks to create im-ages/videos from an adjustable portion of the event data stream. The stacks of space-time coordinates of events are used as inputs and the network is trained to reproduce images based on the spatio-temporal intensity changes. The usefulness of event cameras to generate high dynamic range (HDR) images even in extreme illumination conditions and also non blurred images under rapid motion is also shown. In addition, the possibility of generating very high frame rate videos is demonstrated, theoretically up to 1 million frames per second (FPS) since the temporal resolution of event cameras are about 1 µs. Proposed methods are evaluated by comparing the results with the intensity images captured on the same pixel grid-line of events using online available real datasets and synthetic datasets produced by the event camera simulator.
The temperature of human body is indicative of critical medical information concerning the whole ... more The temperature of human body is indicative of critical medical information concerning the whole body.Abnormal rise in total and regional body temperature is a natural symptom in diagnosing many types of diseases. Thermal imaging (Thermography) utilizes infrared beams, which are fast, non-invasive, andnon-contact. Generally, the output images by this technique are flexible and useful to monitor thetemperature of the human body. The purpose of this study is to determine the diagnostic value of thebreast tissue diseases by the help of thermography.In this paper, we used the non-contact infrared camera INFREC R500 for evaluating the capabilities of thermography. This study was conducted on 60 patients suspected of breast tissue disease, which werereferred to Imam Khomeini Imaging Center. The overall information was obtained from multiplesources including a predefined medical questionnaire, the performed clinical examinations, diagnosticresults obtained from ultrasound images, clinical biopsies and thermography. All of these inputs wereanalyzed by the respective experts of breast cancer. Medical analysis indicated that the use of Thermography as well as the asymmetry technique is useful in identifying hypo echoic and cysticmasses. It should be noted that the patient should not suffer from breast discharge.The Accuracy of the asymmetry identification technique for identifying hypo echoic and cystic masses is91/89% and 92/30%.In the same manner, the accuracy of identifying the exact location is 61/53% and75%. This method is also effective in identifying heterogeneous, fibro-adenoma, and Intra-ductalmasses. But it is unable to identify Iso-echo and calcified mass.According to the results of the investigation, Thermography is useful in the initial screening andsupplementation of diagnostic procedures due to its safety (its non-ionizing radiation property whichacts as everyday immersed heat), low cost and the exact recognition of diseases of the breast tissue/
2010 IEEE International Conference on Computational Intelligence and Computing Research, 2010
... 1Sabzevar Tarbiat Moallem University, Sabzevar, Iran 2Islamic Azad University Najafabad, Esfa... more ... 1Sabzevar Tarbiat Moallem University, Sabzevar, Iran 2Islamic Azad University Najafabad, Esfahan, Iran mostafavi@ieee.org mydigitalemail@gmail.com h_pourghasem@iaun.ac.ir jhaddadnia@sttu.ac.ir ... i(x,y)a , b = 1. Imin if i(x, y) <= a 2. i(x,y) × Imax / (ba) if a < i(x, y) < b 3 ...
2010 IEEE Asia Pacific Conference on Circuits and Systems, 2010
An ultra-low power adjustable triangle wave generator with a multi tunable amplitude and frequenc... more An ultra-low power adjustable triangle wave generator with a multi tunable amplitude and frequency is introduced in this paper. The proposed circuit consists of a Schmitt trigger and a current source. The overall nonlinearity of the TWG circuit is less than 2% in its current-to-frequency transfer characteristic. The tunable frequency and amplitude range are 10KHz to 40KHz and 0.1V-1.7V respectively. The topology is suitable for VLSI realization and can be used in the WINeR system.
2010 IEEE Asia Pacific Conference on Circuits and Systems, 2010
Abstract In this paper a low-power low-noise amplifier for neural recording and biomedical applic... more Abstract In this paper a low-power low-noise amplifier for neural recording and biomedical applications is presented. The frequency band of the amplifier is tunable. It has a gain of 28.3 dB. The low and the high cut-off frequency can be adjusted from 24 mHz to 30.6 Hz ...
2010 International Conference on Computer Applications and Industrial Electronics, 2010
In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters appli... more In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters applied to printed text and handwriting images. CL Filters and their associated Coordinate Logic Operations (CLOs) are widely used in common practical image process applications like noise removing, magnification, opening, closing, skeletonization, coding, edge detection, feature extraction, and fractal modeling. Using coordinate logic filters increases the efficiency, simplicity and more than all the programs speed compared to morphological filters. Unlike other methods like applying Wiener filter or Median filter that are dependent to the noise applied (Gaussian or salt & pepper noise) and to the type of the input image, (handwritten text images or printed text images) coordinate logic filters are not only independent to these variations but also on the other hand accurate and fast.
International Journal of Image, Graphics and Signal Processing, 2013
A challenging field in image processing and computer graphics is to have higher frequency details... more A challenging field in image processing and computer graphics is to have higher frequency details by super resolving facial images. Unlike similar papers in this field, this paper introduces a practical face hallucinating approach with higher quality output images. The image reconstruction was based on a set of high and low resolution image pairs. Each image is divided into defined patches with overlapped regions. A patch from a defined location is removed from the low resolution (LR) input image and is compared with the LR patches of the training images with the same location. Each defined LR patch has a defined high resolution (HR) patch. Based on the Euclidean distance comparison, each patch of every single image in the training images database receives a specific weight. This weight is transferred to its relevant HR patch identically. The sum of the gained weights for one specific location of a patch is equal to unity. The HR output image is constructed by integrating the HR hallucinated patches.
2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019
Event cameras have a lot of advantages over traditional cameras, such as low latency, high tempor... more Event cameras have a lot of advantages over traditional cameras, such as low latency, high temporal resolution, and high dynamic range. However, since the outputs of event cameras are the sequences of asynchronous events over time rather than actual intensity images, existing algorithms could not be directly applied. Therefore, it is demanding to generate intensity images from events for other tasks. In this paper, we unlock the potential of event camera-based conditional generative adversarial networks to create images/videos from an adjustable portion of the event data stream. The stacks of space-time coordinates of events are used as inputs and the network is trained to reproduce images based on the spatio-temporal intensity changes. The usefulness of event cameras to generate high dynamic range (HDR) images even in extreme illumination conditions and also non blurred images under rapid motion is also shown. In addition, the possibility of generating very high frame rate videos is demonstrated, theoretically up to 1 million frames per second (FPS) since the temporal resolution of event cameras are about 1 µs. Proposed methods are evaluated by comparing the results with the intensity images captured on the same pixel grid-line of events using online available real datasets and synthetic datasets produced by the event camera simulator.
An event camera detects per-pixel intensity difference and produces asynchronous event stream wit... more An event camera detects per-pixel intensity difference and produces asynchronous event stream with low latency, high dynamic range, and low power consumption. As a trade-off, the event camera has low spatial resolution. We propose an end-to-end network to reconstruct high resolution, high dynamic range (HDR) images directly from the event stream. We evaluate our algorithm on both simulated and real-world sequences and verify that it captures fine details of a scene and outperforms the combination of the state-of-the-art event to image algorithms with the state-of-the-art super resolution schemes in many quantitative measures by large margins. We further extend our method by using the active sensor pixel (APS) frames or reconstructing images iteratively.
2010 International Conference on Computer Applications and Industrial Electronics, 2010
In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters appli... more In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters applied to printed text and handwriting images. CL Filters and their associated Coordinate Logic Operations (CLOs) are widely used in common practical image process applications like noise removing, magnification, opening, closing, skeletonization, coding, edge detection, feature extraction, and fractal modeling. Using coordinate logic filters increases the efficiency, simplicity and more than all the programs speed compared to morphological filters. Unlike other methods like applying Wiener filter or Median filter that are dependent to the noise applied (Gaussian or salt & pepper noise) and to the type of the input image, (handwritten text images or printed text images) coordinate logic filters are not only independent to these variations but also on the other hand accurate and fast.
2010 International Conference on Computer Applications and Industrial Electronics, 2010
In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters appli... more In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters applied to printed text and handwriting images. CL Filters and their associated Coordinate Logic Operations (CLOs) are widely used in common practical image process applications like noise removing, magnification, opening, closing, skeletonization, coding, edge detection, feature extraction, and fractal modeling. Using coordinate logic filters increases the efficiency, simplicity and more than all the programs speed compared to morphological filters. Unlike other methods like applying Wiener filter or Median filter that are dependent to the noise applied (Gaussian or salt & pepper noise) and to the type of the input image, (handwritten text images or printed text images) coordinate logic filters are not only independent to these variations but also on the other hand accurate and fast.
An event camera detects per-pixel intensity difference and produces asynchronous event stream wit... more An event camera detects per-pixel intensity difference and produces asynchronous event stream with low latency, high dynamic range, and low power consumption. As a trade-off, the event camera has low spatial resolution. We propose an end-to-end network to reconstruct high resolution, high dynamic range (HDR) images directly from the event stream. We evaluate our algorithm on both simulated and real-world sequences and verify that it captures fine details of a scene and outperforms the combination of the state-of-the-art event to image algorithms with the state-of- the-art super resolution schemes in many quantitative measures by large margins. We further extend our method by using the active sensor pixel (APS) frames or reconstructing images iteratively.
Event cameras have a lot of advantages over traditional cameras, such as low latency, high tempor... more Event cameras have a lot of advantages over traditional cameras, such as low latency, high temporal resolution, and high dynamic range. However, since the outputs of event cameras are the sequences of asynchronous events over time rather than actual intensity images, existing algorithms could not be directly applied. Therefore, it is demanding to generate intensity images from events for other tasks. In this paper, we unlock the potential of event camera-based conditional generative adversarial networks to create im-ages/videos from an adjustable portion of the event data stream. The stacks of space-time coordinates of events are used as inputs and the network is trained to reproduce images based on the spatio-temporal intensity changes. The usefulness of event cameras to generate high dynamic range (HDR) images even in extreme illumination conditions and also non blurred images under rapid motion is also shown. In addition, the possibility of generating very high frame rate videos is demonstrated, theoretically up to 1 million frames per second (FPS) since the temporal resolution of event cameras are about 1 µs. Proposed methods are evaluated by comparing the results with the intensity images captured on the same pixel grid-line of events using online available real datasets and synthetic datasets produced by the event camera simulator.
The temperature of human body is indicative of critical medical information concerning the whole ... more The temperature of human body is indicative of critical medical information concerning the whole body.Abnormal rise in total and regional body temperature is a natural symptom in diagnosing many types of diseases. Thermal imaging (Thermography) utilizes infrared beams, which are fast, non-invasive, andnon-contact. Generally, the output images by this technique are flexible and useful to monitor thetemperature of the human body. The purpose of this study is to determine the diagnostic value of thebreast tissue diseases by the help of thermography.In this paper, we used the non-contact infrared camera INFREC R500 for evaluating the capabilities of thermography. This study was conducted on 60 patients suspected of breast tissue disease, which werereferred to Imam Khomeini Imaging Center. The overall information was obtained from multiplesources including a predefined medical questionnaire, the performed clinical examinations, diagnosticresults obtained from ultrasound images, clinical biopsies and thermography. All of these inputs wereanalyzed by the respective experts of breast cancer. Medical analysis indicated that the use of Thermography as well as the asymmetry technique is useful in identifying hypo echoic and cysticmasses. It should be noted that the patient should not suffer from breast discharge.The Accuracy of the asymmetry identification technique for identifying hypo echoic and cystic masses is91/89% and 92/30%.In the same manner, the accuracy of identifying the exact location is 61/53% and75%. This method is also effective in identifying heterogeneous, fibro-adenoma, and Intra-ductalmasses. But it is unable to identify Iso-echo and calcified mass.According to the results of the investigation, Thermography is useful in the initial screening andsupplementation of diagnostic procedures due to its safety (its non-ionizing radiation property whichacts as everyday immersed heat), low cost and the exact recognition of diseases of the breast tissue/
2010 IEEE International Conference on Computational Intelligence and Computing Research, 2010
... 1Sabzevar Tarbiat Moallem University, Sabzevar, Iran 2Islamic Azad University Najafabad, Esfa... more ... 1Sabzevar Tarbiat Moallem University, Sabzevar, Iran 2Islamic Azad University Najafabad, Esfahan, Iran mostafavi@ieee.org mydigitalemail@gmail.com h_pourghasem@iaun.ac.ir jhaddadnia@sttu.ac.ir ... i(x,y)a , b = 1. Imin if i(x, y) <= a 2. i(x,y) × Imax / (ba) if a < i(x, y) < b 3 ...
2010 IEEE Asia Pacific Conference on Circuits and Systems, 2010
An ultra-low power adjustable triangle wave generator with a multi tunable amplitude and frequenc... more An ultra-low power adjustable triangle wave generator with a multi tunable amplitude and frequency is introduced in this paper. The proposed circuit consists of a Schmitt trigger and a current source. The overall nonlinearity of the TWG circuit is less than 2% in its current-to-frequency transfer characteristic. The tunable frequency and amplitude range are 10KHz to 40KHz and 0.1V-1.7V respectively. The topology is suitable for VLSI realization and can be used in the WINeR system.
2010 IEEE Asia Pacific Conference on Circuits and Systems, 2010
Abstract In this paper a low-power low-noise amplifier for neural recording and biomedical applic... more Abstract In this paper a low-power low-noise amplifier for neural recording and biomedical applications is presented. The frequency band of the amplifier is tunable. It has a gain of 28.3 dB. The low and the high cut-off frequency can be adjusted from 24 mHz to 30.6 Hz ...
2010 International Conference on Computer Applications and Industrial Electronics, 2010
In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters appli... more In this paper we present a novel noise reduction method using Coordinate Logic (CL) filters applied to printed text and handwriting images. CL Filters and their associated Coordinate Logic Operations (CLOs) are widely used in common practical image process applications like noise removing, magnification, opening, closing, skeletonization, coding, edge detection, feature extraction, and fractal modeling. Using coordinate logic filters increases the efficiency, simplicity and more than all the programs speed compared to morphological filters. Unlike other methods like applying Wiener filter or Median filter that are dependent to the noise applied (Gaussian or salt & pepper noise) and to the type of the input image, (handwritten text images or printed text images) coordinate logic filters are not only independent to these variations but also on the other hand accurate and fast.
International Journal of Image, Graphics and Signal Processing, 2013
A challenging field in image processing and computer graphics is to have higher frequency details... more A challenging field in image processing and computer graphics is to have higher frequency details by super resolving facial images. Unlike similar papers in this field, this paper introduces a practical face hallucinating approach with higher quality output images. The image reconstruction was based on a set of high and low resolution image pairs. Each image is divided into defined patches with overlapped regions. A patch from a defined location is removed from the low resolution (LR) input image and is compared with the LR patches of the training images with the same location. Each defined LR patch has a defined high resolution (HR) patch. Based on the Euclidean distance comparison, each patch of every single image in the training images database receives a specific weight. This weight is transferred to its relevant HR patch identically. The sum of the gained weights for one specific location of a patch is equal to unity. The HR output image is constructed by integrating the HR hallucinated patches.