-
Learn to Categorize or Categorize to Learn? Self-Coding for Generalized Category Discovery
Authors:
Sarah Rastegar,
Hazel Doughty,
Cees G. M. Snoek
Abstract:
In the quest for unveiling novel categories at test time, we confront the inherent limitations of traditional supervised recognition models that are restricted by a predefined category set. While strides have been made in the realms of self-supervised and open-world learning towards test-time category discovery, a crucial yet often overlooked question persists: what exactly delineates a category?…
▽ More
In the quest for unveiling novel categories at test time, we confront the inherent limitations of traditional supervised recognition models that are restricted by a predefined category set. While strides have been made in the realms of self-supervised and open-world learning towards test-time category discovery, a crucial yet often overlooked question persists: what exactly delineates a category? In this paper, we conceptualize a category through the lens of optimization, viewing it as an optimal solution to a well-defined problem. Harnessing this unique conceptualization, we propose a novel, efficient and self-supervised method capable of discovering previously unknown categories at test time. A salient feature of our approach is the assignment of minimum length category codes to individual data instances, which encapsulates the implicit category hierarchy prevalent in real-world datasets. This mechanism affords us enhanced control over category granularity, thereby equip** our model to handle fine-grained categories adeptly. Experimental evaluations, bolstered by state-of-the-art benchmark comparisons, testify to the efficacy of our solution in managing unknown categories at test time. Furthermore, we fortify our proposition with a theoretical foundation, providing proof of its optimality. Our code is available at https://github.com/SarahRastegar/InfoSieve.
△ Less
Submitted 18 January, 2024; v1 submitted 30 October, 2023;
originally announced October 2023.
-
First Principles Study of Photocatalytic Water Splitting by M$_1$M$_2$CO$_2$ (M$_1$ = Zr,Hf; M$_2$ = Hf,Ti,Sc) MXenes
Authors:
Sima Rastegar,
Alireza Rastkar Ebrahimzadeh,
Jaber Jahanbin Sardroodi
Abstract:
Using density functional theory (DFT), we investigated the structural, electronic and optical properties of functionalized and doped MXenes such as M$_1$M$_2$CO$_2$ (M$_1$ = Zr,Hf; M$_2$ = Hf,Ti,Sc). This study aimed to find a suitable photocatalyst that would work well in the water splitting process. Among the calculated nanostructures, MXenes ZrHfCO$_2$ and ZrTiCO$_2$ were chosen as the suitable…
▽ More
Using density functional theory (DFT), we investigated the structural, electronic and optical properties of functionalized and doped MXenes such as M$_1$M$_2$CO$_2$ (M$_1$ = Zr,Hf; M$_2$ = Hf,Ti,Sc). This study aimed to find a suitable photocatalyst that would work well in the water splitting process. Among the calculated nanostructures, MXenes ZrHfCO$_2$ and ZrTiCO$_2$ were chosen as the suitable photocatalysts for the water splitting process. The calculated value of the band gaps with the GGA-PBE functional was 1.08(0.79) eV for the ZrHfCO$_2$ (ZrTiCO$_2$) monolayer. Also, the band gaps for these monolayers with the HSE06 hybrid functional were 1.86 and 1.57 eV, respectively. These MXenes' optical properties, such as complex dielectric function, refractive index, extinction coefficient, and reflectivity, were also investigated. The results showed that these monolayers had good absorption in the visible and ultraviolet regions. Additionally, we discovered that ZrHfCO$_2$ and ZrTiCO$_2$ MXenes could be used for the water splitting process by calculating the photocatalytic properties. Meanwhile, the results showed that the monolayers of M$_1$M$_2$CO$_2$ could be promising candidates for photocatalytic, solar energy, and optoelectronic applications.
△ Less
Submitted 25 May, 2023;
originally announced May 2023.
-
Spin Hall and Edelstein Effects in pseudospin-1 systems: significant contribution of vertex corrections
Authors:
S. Rastegar,
A. Phirouznia
Abstract:
Edelstein and spin Hall effect response functions for a two-dimensional (2D) system of pseudospin-1 particles is investigated. These two response functions denoted by $ σ_{SH} $ and $ σ_{EE} $ have been analyzed in a pseudospin-1 particle system in the presence of the Rashba-type spin-orbit interaction using the Kubo-Streda technique and vertex corrections with non-magnetic impurities. Then, for a…
▽ More
Edelstein and spin Hall effect response functions for a two-dimensional (2D) system of pseudospin-1 particles is investigated. These two response functions denoted by $ σ_{SH} $ and $ σ_{EE} $ have been analyzed in a pseudospin-1 particle system in the presence of the Rashba-type spin-orbit interaction using the Kubo-Streda technique and vertex corrections with non-magnetic impurities. Then, for a given range of the Rashba coupling, response functions of the spin Hall effect (SHE) and Edelstein effect (EE) have been estimated at various energy gaps and Fermi energies. Results indicate that in this type of the two-dimensional materials, SHE and EE are essentially induced by the vertex corrections and without considering these corrections the amount of these effects are really negligible. It has also been realized that SHE and EE conductivities can be modulated by the Rashba coupling strength at low and intermediate level of this spin-orbit type interaction.
△ Less
Submitted 2 January, 2023;
originally announced January 2023.
-
Thermal Phase Curves of XO-3b: an Eccentric Hot Jupiter at the Deuterium Burning Limit
Authors:
Lisa Dang,
Taylor J. Bell,
Nicolas B. Cowan,
Daniel Thorngren,
Tiffany Kataria,
Heather A. Knutson,
Nikole K. Lewis,
Keivan G. Stassun,
Jonathan J. Fortney,
Eric Agol,
Gregory P. Laughlin,
Adam Burrows,
Karen A. Collins,
Drake Deming,
Diana Jovmir,
Jonathan Langton,
Sara Rastegar,
Adam P. Showman
Abstract:
We report \textit{Spitzer} full-orbit phase observations of the eccentric hot Jupiter XO-3b at 3.6 and 4.5 $μ$m. Our new eclipse depth measurements of $1770 \pm 180$ ppm at 3.6 $μ$m and $1610 \pm 70$ ppm at 4.5 $μ$m show no evidence of the previously reported dayside temperature inversion. We also empirically derive the mass and radius of XO-3b and its host star using Gaia DR3's parallax measureme…
▽ More
We report \textit{Spitzer} full-orbit phase observations of the eccentric hot Jupiter XO-3b at 3.6 and 4.5 $μ$m. Our new eclipse depth measurements of $1770 \pm 180$ ppm at 3.6 $μ$m and $1610 \pm 70$ ppm at 4.5 $μ$m show no evidence of the previously reported dayside temperature inversion. We also empirically derive the mass and radius of XO-3b and its host star using Gaia DR3's parallax measurement and find a planetary mass $M_p=11.79 \pm 0.98 ~M_{\rm{Jup}}$ and radius $R_p=1.295 \pm 0.066 ~R_{\rm{Jup}}$. We compare our \textit{Spitzer} observations with multiple atmospheric models to constrain the radiative and advective properties of XO-3b. While the decorrelated 4.5 $μ$m observations are pristine, the 3.6 $μ$m phase curve remains polluted with detector systematics due to larger amplitude intrapixel sensitivity variations in this channel. We focus our analysis on the more reliable 4.5 $μ$m phase curve and fit an energy balance model with solid body rotation to estimate the zonal wind speed and the pressure of the bottom of the mixed layer. Our energy balance model fit suggests an eastward equatorial wind speed of $3.13 ^{+0.26} _{-0.83}$ km/s, an atmospheric mixed layer down to $2.40 ^{+0.92} _{-0.16}$ bar, and Bond albedo of $0.106 ^{+0.008} _{-0.106}$. We assume that the wind speed and mixed layer depth are constant throughout the orbit. We compare our observations with a 1D planet-averaged model predictions at apoapse and periapse and 3D general circulation model (GCM) predictions for XO-3b. We also investigate the inflated radius of XO-3b and find that it would require an unusually large amount of internal heating to explain the observed planetary radius.
△ Less
Submitted 5 November, 2021;
originally announced November 2021.
-
Delay Analysis of Base Station Flow Table in SDN-enabled Radio Access Networks
Authors:
Seyed Hamed Rastegar,
Aliazam Abbasfar,
Vahid Shah-Mansouri
Abstract:
Future generation wireless networks are designed with extremely low delay requirements which makes even small contributed delays important. On the other hand, software defined networking (SDN) has been introduced as a key enabler of future wireless and cellular networks in order to make them more flexible. In SDN, a central controller manages all network equipments by setting the match-action pair…
▽ More
Future generation wireless networks are designed with extremely low delay requirements which makes even small contributed delays important. On the other hand, software defined networking (SDN) has been introduced as a key enabler of future wireless and cellular networks in order to make them more flexible. In SDN, a central controller manages all network equipments by setting the match-action pairs in flow tables of the devices. However, these flow tables have limited capacity and thus are not capable of storing the rules of all the users. In this paper, we consider an SDN-enabled base station (SD-BS) in a cell equipped with a limited capacity flow table. We analyze the expected delay incurred in processing of the incoming packets to the SD-BS and present a mathematical expression for it in terms of density of the users and cell area.
△ Less
Submitted 12 June, 2021;
originally announced June 2021.
-
Implications of the TTV-Detection of Close-In Terrestrial Planets Around M Stars for Their Origin and Dynamical Evolution
Authors:
Nader Haghighipour,
Sara Rastegar
Abstract:
It has been shown that an Earth-size planet or a super-Earth, in resonance with a transiting Jupiter-like body in a short-period orbit around an M star, can create detectable TTV signals (Kirste \& Haghighipour, 2011). Given the low masses of M stars and their circumstellar disks, it is expected that such a transiting giant planet to have formed at large distances and migrated to its close-in orbi…
▽ More
It has been shown that an Earth-size planet or a super-Earth, in resonance with a transiting Jupiter-like body in a short-period orbit around an M star, can create detectable TTV signals (Kirste \& Haghighipour, 2011). Given the low masses of M stars and their circumstellar disks, it is expected that such a transiting giant planet to have formed at large distances and migrated to its close-in orbit. That implies, if such systems are discovered around M stars, the terrestrial planet had to form during the migration of the giant planet. The formation of this object may be either in-situ (in a close-in orbit) followed by its capture in resonance, or the object is formed at larger distances where it was subsequently captured in a resonance with the migrating giant planet. We have investigated these two scenarios by simulating the dynamics of a disk of protoplanetary embryos and the formation of terrestrial planets during the migration of a Jupiter-like planet around an M star. Results suggest that unless the migration of the giant planet is very slow (slower than 1E-7 AU/year), it is unlikely that the close-in terrestrial planet is formed in-situ. If a terrestrial planet is detected in a mean-motion resonance with a close-in giant planet around an M star, the terrestrial planet was most likley formed at large distances and carried to its close-in resonant orbit by the migrating giant body.
△ Less
Submitted 10 December, 2010;
originally announced December 2010.
