Ramesh Neelamani earned his PhD at Rice University (2003) and is currently employed at ExxonMobil Production Company. His principal impact to date has been in practical applications of sparsity theory and curvelets (e.g., to noise attenuation, multiple attenuation), multiscale multimedia signal and image processing, inverse halftoning, and blind estimation of the color space in which an image has been compressed. Ramesh won second prize in the IEEE all-India student paper contest (1995). His long and diverse list of patents and publications is very impressive. Quoting from one of Neelamani’s nomination letters: “His 2008 paper ‘Coherent and random noise attenuation using the curvelet transform’ was a pioneering application of curvelets to noise attenuation in seismic data and contains impressive fi eld data examples. His 2010 paper ‘Adaptive subtraction using complex-valued curvelet transforms’ is an important contribution in the fi eld of multiple attenuation and introduces a defi nition of the phase of a complex-valued curvelet transform and describes a new adaptive subtraction method which go a long way towards solving the long-standing problem of attenuating multiples while preserving primaries. Ramesh’s most recent GEOPHYSICS paper, published in December 2010, will have a signifi cant impact in improving effi ciency of forward modeling, full waveform inversion and seismic acquisition with simultaneous sources.” Ramesh has published four articles in GEOPHYSICS, one in THE LEADING EDGE, and six expanded abstracts. He is the fi rst author on six of these SEG publications.
Biography Citation for the J. Clarence Karcher Award
Ramesh (Neelsh) Neelamani was introduced to seismic data for the fi rst time in 2003 when he joined ExxonMobil Upstream Research Company, and today he is being awarded the J. Clarence Karcher award for significant contributions to exploration geophysics. The large number of his publications and their diverse topic areas are as impressive as the number of different collaborations they represent. We are honored to write this citation as we are just three of dozens of geophysicists who have collaborated with Neelsh, as he his known to friends. Neelsh has contributed to diverse fields of multiple and noise attenuation, forward modeling, interpolation and regularization, full wavefi eld inversion, spectral decomposition, and seismic acquisition. He has published 16 journal and book articles, including a chapter on deconvolution, in the Springer Handbook of Signal Processing in Acoustics, 9 patent applications, and 29 conference publications. He has continued collaborating and publishing since he moved in 2008 from research to assignments in exploration and then production.
Neelsh’s academic training was not in geophysics but in electrical engineering. During his PhD studies at Rice University, he focused on inverse problems that arise in image processing and formulated novel approaches that exploited the structure of the desired solution to overcome the solution’s nonuniqueness and the presence of noise in the observed data. His thesis work has been well-cited not only in image processing, but also in geophysics, medical imaging, cryptography, discrete mathematics, and engineering. One of his thesis papers (“ForWaRD: Fourier-wavelet regularized deconvolution for ill-conditioned systems”) was published in the IEEE Transactions on Signal Processing in 2002 has received more than 200 citations.
Neelsh was among the first to apply the emerging theory of compressed sensing to gain key insights to simultaneous sourcing. His presentation on this topic was ranked among the top 30 papers at SEG’s 2008 Annual Meeting. Simultaneous sourcing is currently receiving much attention with its potential to increase the efficiency of seismic acquisition and forward modeling for full waveform inversion. For the forward modeling problem, Neelsh demonstrated low-noise separation, not of dozens of simultaneous sources, but of 8192 simultaneous sources shooting into a single receiver. In a 2010 GEOPHYSICS paper, he analyzed the limitations for separation including the amount of input data required to achieve the desired output signal-to-interference- noise quality.
Neelsh was also an early advocate of using the curvelet transform for noise suppression and multiple attenuation. The curvelet transform can capture the directionality, localization, and varying frequency content of features in data. Neelsh demonstrated that in the curvelet transform domain, signal and noise components of seismic data have minimal overlap, and can be easily separated. He developed a curvelet-based noise attenuation method and applied it to a noisy 3D seismic cube from a carbonate environment (THE LEADING EDGE, 2008). He further introduced the concept of a complex-valued curvelet transform and devised a novel adaptive-subtraction method for multiple attenuation (GEOPHYSICS, 2010).
Neelsh’s enthusiasm, talent and ability in solving challenging problems and willingness to work with and mentor others unquestionably put him in the rank of young geophysicists with extraordinary ability.