arXiv subjects

Suman Saha

Publications and source records attributed to Suman Saha.

At least 55 records · Page 3Linked to original sources

arXiv2025

Next-Gen Education: Enhancing AI for Microlearning

This paper explores integrating microlearning strategies into university curricula, particularly in computer science education, to counteract the decline in class attendance and engagement in US universities after COVID. As students increasingly opt for remote learning and recorded lectures, traditional educational approaches struggle to maintain engagement and effectiveness. Microlearning, which breaks complex subjects into manageable units, is proposed to address shorter attention spans and enhance educational outcomes. It uses interactive formats such as videos, quizzes, flashcards, and scenario-based exercises, which are especially beneficial for topics like algorithms and programming logic requiring deep understanding and ongoing practice. Adoption of microlearning is often limited by the effort needed to create such materials. This paper proposes leveraging AI tools, specifically ChatGPT, to reduce the workload for educators by automating the creation of supplementary materials. While AI can automate certain tasks, educators remain essential in guiding and shaping the learning process. This AI-enhanced approach ensures course content is kept current with the latest research and technology, with educators providing context and insights. By examining AI capabilities in microlearning, this study shows the potential to transform educational practices and outcomes in computer science, offering a practical model for combining advanced technology with established teaching methods.

cs.CY↗

arXiv2025

Universal Patterns in the Blockchain: Analysis of EOAs and Smart Contracts in ERC20 Token Networks

Scaling laws offer a powerful lens to understand complex transactional behaviors in decentralized systems. This study reveals distinctive statistical signatures in the transactional dynamics of ERC20 tokens on the Ethereum blockchain by examining over 44 million token transfers between July 2017 and March 2018 (9-month period). Transactions are categorized into four types: EOA--EOA, EOA--SC, SC-EOA, and SC-SC based on whether the interacting addresses are Externally Owned Accounts (EOAs) or Smart Contracts (SCs), and analyzed across three equal periods (each of 3 months). To identify universal statistical patterns, we investigate the presence of two canonical scaling laws: power law distributions and temporal Taylor's law (TL). EOA-driven transactions exhibit consistent statistical behavior, including a near-linear relationship between trade volume and unique partners with stable power law exponents ($γ\approx 2.3$), and adherence to TL with scaling coefficients ($β\approx 2.3$). In contrast, interactions involving SCs, especially SC-SC, exhibit sublinear scaling, unstable power-law exponents, and significantly fluctuating Taylor coefficients (variation in $β$ to be $Δβ= 0.51$). Moreover, SC-driven activity displays heavier-tailed distributions ($γ< 2$), indicating bursty and algorithm-driven activity. These findings reveal the characteristic differences between human-controlled and automated transaction behaviors in blockchain ecosystems. By uncovering universal scaling behaviors through the integration of complex systems theory and blockchain data analytics, this work provides a principled framework for understanding the underlying mechanisms of decentralized financial systems.

q-fin.ST↗

arXiv2025

Giant Outer Transiting Exoplanet Mass (GOT 'EM) Survey. VI: Confirmation of a Long-Period Giant Planet Discovered with a Single TESS Transit

We report the discovery and confirmation of TOI-4465 b, a $1.25^{+0.08}_{-0.07}~R_{J}$, $5.89\pm0.26~M_{J}$ giant planet orbiting a G dwarf star at $d\simeq$ 122 pc. The planet was detected as a single-transit event in data from Sector 40 of the Transiting Exoplanet Survey Satellite (TESS) mission. Radial velocity (RV) observations of TOI-4465 showed a planetary signal with an orbital period of $\sim$102 days, and an orbital eccentricity of $e=0.24\pm0.01$. TESS re-observed TOI-4465 in Sector 53 and Sector 80, but did not detect another transit of TOI-4465 b, as the planet was not expected to transit during these observations based on the RV period. A global ground-based photometry campaign was initiated to observe another transit of TOI-4465 b after the RV period determination. The $\sim$12 hour-long transit event was captured from multiple sites around the world, and included observations from 24 citizen scientists, confirming the orbital period as $\sim$102 days. TOI-4465 b is a relatively dense ($3.73\pm0.53~\rm{g/cm^3}$), temperate (375-478 K) giant planet. Based on giant planet structure models, TOI-4465 b appears to be enriched in heavy elements at a level consistent with late-stage accretion of icy planetesimals. Additionally, we explore TOI-4465 b's potential for atmospheric characterization, and obliquity measurement. Increasing the number of long-period planets by confirming single-transit events is crucial for understanding the frequency and demographics of planet populations in the outer regions of planetary systems.

astro-ph.EP↗

arXiv2025

A High Geometric Albedo for LTT9779b Points Towards a Metal-rich Atmosphere and Silicate Clouds

Aims: In this work, we aim to confirm the high albedo of the benchmark ultrahot Neptune LTT9779b using 20 secondary eclipse measurements of the planet observed with CHEOPS. In addition, we perform a search for variability in the reflected light intensity of the planet as a function of time. Methods: First, we used the TESS follow-up data of LTT9779b from three sectors (2, 29, and 69) to remodel the transit signature and estimate an updated set of transit and ephemeris parameters, which were directly used in the modeling of the secondary eclipse lightcurves. This involved a critical noise-treatment algorithm, including sophisticated techniques such as wavelet denoising and Gaussian Process (GP) regression, to constrain noise levels from various sources. In addition to using the officially released reduced aperture photometry data from CHEOPS DRP, we also reduced the raw data using an independent PSF photometry pipeline, known as PIPE, to verify the robustness of our analysis. The extracted secondary eclipse lightcurves were modeled using the PYCHEOPS package, where we have detrended the background noise correlated with the spacecraft roll angle, originating from the inhomogeneous and asymmetric shape of the CHEOPS point spread function, using an N-order glint function. Results: Our independent lightcurve analyses have resulted in consistent estimations of the eclipse depths, with values of 89.9$\pm$13.7 ppm for the DRP analysis and 85.2$\pm$13.1 ppm from PIPE, indicating a high degree of statistical agreement. Adopting the DRP value yields a highly constrained geometric albedo of 0.73$\pm$0.11. No significant eclipse depth variability is detected down to a level of $\sim$37 ppm. Conclusions: Our results confirm that LTT9779b exhibits a strikingly high optical albedo, which substantially reduces the internal energy budget of the planet compared to more opaque...

astro-ph.EP↗

arXiv2025

Leveraging LLM to Strengthen ML-Based Cross-Site Scripting Detection

According to the Open Web Application Security Project (OWASP), Cross-Site Scripting (XSS) is a critical security vulnerability. Despite decades of research, XSS remains among the top 10 security vulnerabilities. Researchers have proposed various techniques to protect systems from XSS attacks, with machine learning (ML) being one of the most widely used methods. An ML model is trained on a dataset to identify potential XSS threats, making its effectiveness highly dependent on the size and diversity of the training data. A variation of XSS is obfuscated XSS, where attackers apply obfuscation techniques to alter the code's structure, making it challenging for security systems to detect its malicious intent. Our study's random forest model was trained on traditional (non-obfuscated) XSS data achieved 99.8% accuracy. However, when tested against obfuscated XSS samples, accuracy dropped to 81.9%, underscoring the importance of training ML models with obfuscated data to improve their effectiveness in detecting XSS attacks. A significant challenge is to generate highly complex obfuscated code despite the availability of several public tools. These tools can only produce obfuscation up to certain levels of complexity. In our proposed system, we fine-tune a Large Language Model (LLM) to generate complex obfuscated XSS payloads automatically. By transforming original XSS samples into diverse obfuscated variants, we create challenging training data for ML model evaluation. Our approach achieved a 99.5% accuracy rate with the obfuscated dataset. We also found that the obfuscated samples generated by the LLMs were 28.1% more complex than those created by other tools, significantly improving the model's ability to handle advanced XSS attacks and making it more effective for real-world application security.

cs.CR↗

arXiv2025

Updating the Ephemeris and Physical Properties of Five Long-period Transiting Exoplanets Using TESS and CHEOPS

The transiting long-period exoplanets are the most interesting follow-up targets for the next-generation instruments, using the most sophisticated observational techniques. However, the scarcity in their transit events often leads to larger uncertainties in their known ephemeris, also resulting in larger biases in their known physical properties in many cases. In this work, I have used the newer publicly available observations from TESS and CHEOPS for five very interesting long-period transiting exoplanets, i.e., HD95338 b, TOI-2134 c, K2-290 c, TOI-1898 b, and TOI-813 b, combined with the previously reported observations, to reanalyze their transit properties and estimate the updated ephemeris. The analyses also incorporated a critical noise treatment algorithm, which uses well-tested techniques such as wavelet denoising and Gaussian process regression, to effectively reduce the impact of various noise components present in the lightcurves on the estimated parameters. The study has resulted in a more precise estimation of ephemeris for all the targets, with the precision in the estimated periods being better than 5 seconds, except for TOI-813 b, for which the precision in the estimated period is better than 21 seconds. The other transit parameters also got updated, with statistically significant improvements seen in most of the cases, the major notable improvements being in the estimated value of the impact parameter of TOI-1898 b, the orbital semi-major axis of TOI-2134 c, and the radius of HD95338 b. Although long-period exoplanets are expected to show more significant transit timing variations in the presence of other undetected planetary mass objects, no such variations were recorded for these targets.

astro-ph.EP↗

arXiv2025

NGTS-EB-7, an eccentric, long-period, low-mass eclipsing binary

Despite being the most common types of stars in the Galaxy, the physical properties of late M dwarfs are often poorly constrained. A trend of radius inflation compared to evolutionary models has been observed for earlier type M dwarfs in eclipsing binaries, possibly caused by magnetic activity. It is currently unclear whether this trend also extends to later type M dwarfs below the convective boundary. This makes the discovery of lower-mass, fully convective, M dwarfs in eclipsing binaries valuable for testing evolutionary models especially in longer-period binaries where tidal interaction between the primary and secondary is negligible. With this context, we present the discovery of the NGTS-EB-7 AB system, an eclipsing binary containing a late M dwarf secondary and an evolved G-type primary star. The secondary star has a radius of $0.125 \pm 0.006 R_\odot$ , a mass of $0.096 \pm 0.004 M_\odot$ and follows a highly eccentric $(e=0.71436 \pm 0.00085)$ orbit every $193.35875 \pm 0.00034$ days. This makes NGTS-EB-7 AB the third longest-period eclipsing binary system with a secondary smaller than $200 M_J$ with the mass and radius constrained to better than $5 \%$. In addition, NGTS-EB-7 is situated near the centre of the proposed LOPS2 southern field of the upcoming PLATO mission, allowing for detection of the secondary eclipse and measurement of the companion`s temperature. With its long-period and well-constrained physical properties - NGTS-EB-7 B will make a valuable addition to the sample of M dwarfs in eclipsing binaries and help in determining accurate empirical mass/radius relations for later M dwarf stars.

astro-ph.SR↗

arXiv2024

Language-Guided Instance-Aware Domain-Adaptive Panoptic Segmentation

The increasing relevance of panoptic segmentation is tied to the advancements in autonomous driving and AR/VR applications. However, the deployment of such models has been limited due to the expensive nature of dense data annotation, giving rise to unsupervised domain adaptation (UDA). A key challenge in panoptic UDA is reducing the domain gap between a labeled source and an unlabeled target domain while harmonizing the subtasks of semantic and instance segmentation to limit catastrophic interference. While considerable progress has been achieved, existing approaches mainly focus on the adaptation of semantic segmentation. In this work, we focus on incorporating instance-level adaptation via a novel instance-aware cross-domain mixing strategy IMix. IMix significantly enhances the panoptic quality by improving instance segmentation performance. Specifically, we propose inserting high-confidence predicted instances from the target domain onto source images, retaining the exhaustiveness of the resulting pseudo-labels while reducing the injected confirmation bias. Nevertheless, such an enhancement comes at the cost of degraded semantic performance, attributed to catastrophic forgetting. To mitigate this issue, we regularize our semantic branch by employing CLIP-based domain alignment (CDA), exploiting the domain-robustness of natural language prompts. Finally, we present an end-to-end model incorporating these two mechanisms called LIDAPS, achieving state-of-the-art results on all popular panoptic UDA benchmarks.

cs.CV↗

arXiv2024

GEM: A Generalizable Ego-Vision Multimodal World Model for Fine-Grained Ego-Motion, Object Dynamics, and Scene Composition Control

We present GEM, a Generalizable Ego-vision Multimodal world model that predicts future frames using a reference frame, sparse features, human poses, and ego-trajectories. Hence, our model has precise control over object dynamics, ego-agent motion and human poses. GEM generates paired RGB and depth outputs for richer spatial understanding. We introduce autoregressive noise schedules to enable stable long-horizon generations. Our dataset is comprised of 4000+ hours of multimodal data across domains like autonomous driving, egocentric human activities, and drone flights. Pseudo-labels are used to get depth maps, ego-trajectories, and human poses. We use a comprehensive evaluation framework, including a new Control of Object Manipulation (COM) metric, to assess controllability. Experiments show GEM excels at generating diverse, controllable scenarios and temporal consistency over long generations. Code, models, and datasets are fully open-sourced.

cs.CV↗

arXiv2024

NGTS-33b: A Young Super-Jupiter Hosted by a Fast Rotating Massive Hot Star

In the last few decades planet search surveys have been focusing on solar type stars, and only recently the high-mass regimes. This is mostly due to challenges arising from the lack of instrumental precision, and more importantly, the inherent active nature of fast rotating massive stars. Here we report NGTS-33b (TOI-6442b), a super-Jupiter planet with mass, radius and orbital period of 3.6 $\pm$ 0.3 M$_{\rm jup}$, 1.64 $\pm$ 0.07 R$_{\rm jup}$ and $2.827972 \pm 0.000001$ days, respectively. The host is a fast rotating ($0.6654 \pm 0.0006$ day) and hot (T$_{\rm eff}$ = 7437 $\pm$ 72 K) A9V type star, with a mass and radius of 1.60 $\pm$ 0.11 M$_{\odot}$ and 1.47 $\pm$ 0.06 R$_{\odot}$, respectively. Planet structure and Gyrochronology models shows that NGTS-33 is also very young with age limits of 10-50 Myr. In addition, membership analysis points towards the star being part of the Vela OB2 association, which has an age of $\sim$ 20-35 Myr, thus providing further evidences about the young nature of NGTS-33. Its low bulk density of 0.19$\pm$0.03 g cm$^{-3}$ is 13$\%$ smaller than expected when compared to transiting hot Jupiters with similar masses. Such cannot be solely explained by its age, where an up to 15$\%$ inflated atmosphere is expected from planet structure models. Finally, we found that its emission spectroscopy metric is similar to JWST community targets, making the planet an interesting target for atmospheric follow-up. Therefore, NGTS-33b's discovery will not only add to the scarce population of young, massive and hot Jupiters, but will also help place further strong constraints on current formation and evolution models for such planetary systems.

astro-ph.EP↗

arXiv2024

TOI-2490b- The most eccentric brown dwarf transiting in the brown dwarf desert

We report the discovery of the most eccentric transiting brown dwarf in the brown dwarf desert, TOI02490b. The brown dwarf desert is the lack of brown dwarfs around main sequence stars within $\sim3$~AU and is thought to be caused by differences in formation mechanisms between a star and planet. To date, only $\sim40$ transiting brown dwarfs have been confirmed. \systemt is a $73.6\pm2.4$ \mjupnospace, $1.00\pm0.02$ \rjup brown dwarf orbiting a $1.004_{-0.022}^{+0.031}$ \msunnospace, $1.105_{-0.012}^{+0.012}$ \rsun sun-like star on a 60.33~d orbit with an eccentricity of $0.77989\pm0.00049$. The discovery was detected within \tess sectors 5 (30 minute cadence) and 32 (2 minute and 20 second cadence). It was then confirmed with 31 radial velocity measurements with \feros by the WINE collaboration and photometric observations with the Next Generation Transit Survey. Stellar modelling of the host star estimates an age of $\sim8$~Gyr, which is supported by estimations from kinematics likely placing the object within the thin disc. However, this is not consistent with model brown dwarf isochrones for the system age suggesting an inflated radius. Only one other transiting brown dwarf with an eccentricity higher than 0.6 is currently known in the brown dwarf desert. Demographic studies of brown dwarfs have suggested such high eccentricity is indicative of stellar formation mechanisms.

astro-ph.EP↗

arXiv2024

Precise Transit Photometry Using TESS II: Revisiting 28 Additional Transiting Systems With Updated Physical Properties

Precise physical properties of the known transiting exoplanets are essential for their precise atmospheric characterization using modern and upcoming instruments. Leveraging the large volume of high SNR photometric follow-up data from TESS, highly precise physical properties can be estimated for these systems, especially for those discovered using ground-based instruments prior to the TESS mission. In this work, I have used the publicly available TESS follow-up data for 28 transiting systems with 10 $<$ V$_{mag}$ $<$ 10.5, with an aim to update their known physical properties. The observed lightcurves have been analysed by implementing a state-of-the-art critical noise treatment algorithm to effectively reduce both time-correlated and un-correlated noise components, using sophisticated techniques like wavelet denoising and Gaussian-process regression. Compared with the previous studies, the estimated transit parameters are found to be more precise for most of the targets, including a few cases where a larger space-based instrument like Spitzer, Kepler or CHEOPS has been used in the previous study. The large volume of transit observations used for each target has also resulted in a more accurate estimation of the physical properties, as this overcomes any error in parameter estimations from bias present in a smaller volume of data. Thus, comparing with the literature values, statistically significant improvements in the known physical properties of several targeted systems have been reported from this work. The large volume of transit timing information from the analyses was also used to search for Transit Timing Variation trends in these targets, which has resulted in no significant detection.

astro-ph.EP↗

arXiv2024

Four Ways to Improve Verbo-visual Fusion for Dense 3D Visual Grounding

3D visual grounding is the task of localizing the object in a 3D scene which is referred by a description in natural language. With a wide range of applications ranging from autonomous indoor robotics to AR/VR, the task has recently risen in popularity. A common formulation to tackle 3D visual grounding is grounding-by-detection, where localization is done via bounding boxes. However, for real-life applications that require physical interactions, a bounding box insufficiently describes the geometry of an object. We therefore tackle the problem of dense 3D visual grounding, i.e. referral-based 3D instance segmentation. We propose a dense 3D grounding network ConcreteNet, featuring four novel stand-alone modules that aim to improve grounding performance for challenging repetitive instances, i.e. instances with distractors of the same semantic class. First, we introduce a bottom-up attentive fusion module that aims to disambiguate inter-instance relational cues, next, we construct a contrastive training scheme to induce separation in the latent space, we then resolve view-dependent utterances via a learned global camera token, and finally we employ multi-view ensembling to improve referred mask quality. ConcreteNet ranks 1st on the challenging ScanRefer online benchmark and has won the ICCV 3rd Workshop on Language for 3D Scenes "3D Object Localization" challenge. Our code is available at ouenal.github.io/concretenet/.

cs.CV↗

arXiv2024

Planet Hunters NGTS: New Planet Candidates from a Citizen Science Search of the Next Generation Transit Survey Public Data

We present the results from the first two years of the Planet Hunters NGTS citizen science project, which searches for transiting planet candidates in data from the Next Generation Transit Survey (NGTS) by enlisting the help of members of the general public. Over 8,000 registered volunteers reviewed 138,198 light curves from the NGTS Public Data Releases 1 and 2. We utilize a user weighting scheme to combine the classifications of multiple users to identify the most promising planet candidates not initially discovered by the NGTS team. We highlight the five most interesting planet candidates detected through this search, which are all candidate short-period giant planets. This includes the TIC-165227846 system that, if confirmed, would be the lowest-mass star to host a close-in giant planet. We assess the detection efficiency of the project by determining the number of confirmed planets from the NASA Exoplanet Archive and TESS Objects of Interest (TOIs) successfully recovered by this search and find that 74% of confirmed planets and 63% of TOIs detected by NGTS are recovered by the Planet Hunters NGTS project. The identification of new planet candidates shows that the citizen science approach can provide a complementary method to the detection of exoplanets with ground-based surveys such as NGTS.

astro-ph.EP↗

arXiv2024

CS-Mixer: A Cross-Scale Vision MLP Model with Spatial-Channel Mixing

Despite their simpler information fusion designs compared with Vision Transformers and Convolutional Neural Networks, Vision MLP architectures have demonstrated strong performance and high data efficiency in recent research. However, existing works such as CycleMLP and Vision Permutator typically model spatial information in equal-size spatial regions and do not consider cross-scale spatial interactions. Further, their token mixers only model 1- or 2-axis correlations, avoiding 3-axis spatial-channel mixing due to its computational demands. We therefore propose CS-Mixer, a hierarchical Vision MLP that learns dynamic low-rank transformations for spatial-channel mixing through cross-scale local and global aggregation. The proposed methodology achieves competitive results on popular image recognition benchmarks without incurring substantially more compute. Our largest model, CS-Mixer-L, reaches 83.2% top-1 accuracy on ImageNet-1k with 13.7 GFLOPs and 94 M parameters.

cs.CV↗

arXiv2023

EDAPS: Enhanced Domain-Adaptive Panoptic Segmentation

With autonomous industries on the rise, domain adaptation of the visual perception stack is an important research direction due to the cost savings promise. Much prior art was dedicated to domain-adaptive semantic segmentation in the synthetic-to-real context. Despite being a crucial output of the perception stack, panoptic segmentation has been largely overlooked by the domain adaptation community. Therefore, we revisit well-performing domain adaptation strategies from other fields, adapt them to panoptic segmentation, and show that they can effectively enhance panoptic domain adaptation. Further, we study the panoptic network design and propose a novel architecture (EDAPS) designed explicitly for domain-adaptive panoptic segmentation. It uses a shared, domain-robust transformer encoder to facilitate the joint adaptation of semantic and instance features, but task-specific decoders tailored for the specific requirements of both domain-adaptive semantic and instance segmentation. As a result, the performance gap seen in challenging panoptic benchmarks is substantially narrowed. EDAPS significantly improves the state-of-the-art performance for panoptic segmentation UDA by a large margin of 20% on SYNTHIA-to-Cityscapes and even 72% on the more challenging SYNTHIA-to-Mapillary Vistas. The implementation is available at https://github.com/susaha/edaps.

cs.CV↗

arXiv2023

Precise Transit Photometry Using TESS: Updated Physical Properties for 28 Exoplanets Around Bright Stars

The TESS follow-up of a large number of known transiting exoplanets provide unique opportunity to study their physical properties more precisely. Being a space-based telescope, the TESS observations are devoid of any noise component resulting from the interference of Earth's atmosphere. TESS also provides a better probability to observe subsequent transit events owing to its longer uninterrupted time-series observations compared to the ground-based telescopes. Especially, for the exoplanets around bright host-stars, TESS time-series observations provides high SNR lightcurves, which can be used for higher precision studies for these exoplanets. In this work, I have studied the TESS transit photometric follow-up observations of 28 exoplanets around bright stars with $V_{mag}\le$10. The already high SNR lightcurves from TESS have been further processed with a critical noise treatment algorithm, using the wavelet denoising and the Gaussian-process regression techniques, to effectively reduce the noise components both correlated and uncorrelated in time, which were then used to estimate the physical properties of these exoplanets. The study has resulted in very precise values for the physical properties of the target exoplanets, with the improvements in precision being significant for most of the cases compared to the previous studies. Also, since a comparatively large number of transit lightcurves from TESS observations were used to estimate these physical properties for each of the target exoplanets, which removes any bias due to the lack of sufficient datasets, these updated physical properties can be considered extremely accurate and reliable for future studies.

astro-ph.EP↗

arXiv2022

Spatio-Temporal Action Detection Under Large Motion

Current methods for spatiotemporal action tube detection often extend a bounding box proposal at a given keyframe into a 3D temporal cuboid and pool features from nearby frames. However, such pooling fails to accumulate meaningful spatiotemporal features if the position or shape of the actor shows large 2D motion and variability through the frames, due to large camera motion, large actor shape deformation, fast actor action and so on. In this work, we aim to study the performance of cuboid-aware feature aggregation in action detection under large action. Further, we propose to enhance actor feature representation under large motion by tracking actors and performing temporal feature aggregation along the respective tracks. We define the actor motion with intersection-over-union (IoU) between the boxes of action tubes/tracks at various fixed time scales. The action having a large motion would result in lower IoU over time, and slower actions would maintain higher IoU. We find that track-aware feature aggregation consistently achieves a large improvement in action detection performance, especially for actions under large motion compared to the cuboid-aware baseline. As a result, we also report state-of-the-art on the large-scale MultiSports dataset. The Code is available at https://github.com/gurkirt/ActionTrackDetectron.

cs.CV↗