Institution profile

Sharif University of Technology

Academic institutionasia · ir
Official website
Research library290linked papers
Opportunities0open roles
Selected work

Representative Papers

Enhancing Reliability of STT-MRAM Caches by Eliminating Read Disturbance Accumulation

Mar 01, 2019Design, Automation and Test in Europe

This work addresses the critical reliability threat posed by accumulated read-disturb errors in STT-MRAM caches when unverified ECC-protected cache blocks are read in parallel. For the first time, this study formally characterizes the cumulative nature of such errors and proposes REAP-cache, a lightweight mechanism that completely eliminates read-disturb errors without compromising performance. By refining the ECC verification process, optimizing parallel read scheduling, and incorporating a proactive error prevention strategy, REAP-cache dramatically enhances system reliability—improving the cache’s mean time to failure (MTTF) by 171×—while incurring less than 1% area overhead and only a 2.7% energy penalty.

24 citations4 influentialRead paper

ROBIN: Incremental Oblique Interleaved ECC for Reliability Improvement in STT-MRAM Caches

Jan 21, 2019Asia and South Pacific Design Automation Conference

This work addresses the inefficacy of conventional error-correcting codes (ECCs) in STT-MRAM caches, where data-dependent error patterns significantly exacerbate failure rates. To overcome the limitations of traditional ECCs under spatially and temporally correlated errors, the authors propose ROBIN—a novel incremental diagonal-interleaved ECC architecture. ROBIN leverages insights into the data dependency characteristics of error patterns to tailor its encoding structure, thereby transcending the error-correction constraints of conventional schemes. Experimental results demonstrate that ROBIN reduces cache error rates by an average of 28.6× with low overhead, effectively neutralizing the 151.7% error rate increase induced by data-dependent effects and substantially enhancing cache reliability.

20 citationsRead paper

Scanning Trojaned Models Using Out-of-Distribution Samples

Jan 28, 2025

This paper addresses the challenging problem of backdoor (trojan) detection in deep neural networks—particularly in adversarially trained models and zero-training-data scenarios. We propose TRODO, a general-purpose, prior-free scanning method that requires neither knowledge of attack strategies nor access to training data or label mappings. Our core insight is to exploit out-of-distribution (OOD) samples: by generating adversarial perturbations on OOD inputs, we induce “blind spots”—abnormal model behaviors where OOD samples are erroneously classified as in-distribution (ID) with high confidence. TRODO models these blind spots via confidence-score analysis and statistical significance testing to identify backdoors. To our knowledge, TRODO establishes the first blind-spot detection paradigm grounded in OOD adversarial shift. Evaluated across diverse architectures, datasets, and trojan variants—including those embedded in adversarially trained models—TRODO achieves an average detection accuracy exceeding 94%, operates without any training data, and demonstrates strong generalization and deployment robustness.

2 citationsRead paper

RODEO: Robust Outlier Detection via Exposing Adaptive Out-of-Distribution Samples

Jan 28, 2025International Conference on Machine Learning

To address weak generalization and insufficient robustness of image anomaly detection under unknown and challenging anomaly scenarios, this paper proposes a data-driven robust anomaly detection framework. Methodologically, it introduces: (1) a novel text-image cross-modal guided adaptive anomaly exposure mechanism, overcoming limitations of conventional open-set recognition methods in distribution coverage and semantic plausibility; (2) a controllable out-of-distribution (OOD) sample synthesis pipeline built upon Stable Diffusion, integrating adversarial training and feature-space constraints to generate high-quality OOD samples with diversity, conceptual separability, and distributional proximity; and (3) end-to-end co-optimization of the generator and detector. Evaluated on multiple benchmarks, the framework achieves up to 12.7% AUC improvement, significantly enhancing detection rates for unknown-class anomalies while maintaining low false-positive rates. Qualitative analysis confirms that generated samples exhibit clear semantics, visual fidelity, and boundary sensitivity.

2 citationsRead paper

AIMA at SemEval-2024 Task 10: History-Based Emotion Recognition in Hindi-English Code-Mixed Conversations

Jan 19, 2025International Workshop on Semantic Evaluation

This work addresses emotion recognition in Hindi-English code-mixed (Hinglish) dialogues. We propose a history-aware multimodal framework comprising two key components: (1) a Hinglish-to-English translation pre-processing pipeline for linguistic normalization, and (2) a joint contextual modeling architecture integrating bidirectional LSTM or Transformer-based context encoders with an ensemble of multilingual pretrained language models (BERT, RoBERTa, XLM-R). Crucially, we introduce the novel concept of “history-aware contextual modeling”, synergistically coupled with code-mixed translation pre-processing. This design enhances cross-lingual robustness—particularly critical in low-resource emotion recognition in conversations (ERC). Evaluated on SemEval-2024 Task 10 Subtask 1, our approach outperforms all baseline systems, demonstrating the efficacy of jointly leveraging contextual awareness and language normalization for code-mixed emotion classification.

2 citationsRead paper
Recent publications

Latest Papers