MJManju Lata Joshi

Manju Lata Joshi

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Manju Lata Joshi has a total of 6 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Deep Learning Approaches for Image-Based Food Quality Inspection

    by Manju Lata Joshi

    The growing need for safe, hygienic, high-quality food at affordable costs has increased the adoption of deep learning based automated food quality inspection. Often traditional inspection methods are subjective, labor intensive and not suitable for real time applications. In this chapter, we cover several key deep learning methods as they apply to food image analysis, particularly CNNs, transfer learning, GANs such as WGAN and LSTMs and an introduction to Vision Transformers. These methods can confer abilities for defect detection, freshness evaluation, contamination monitoring, ripeness assessment and quality grading of fruits/vegetables/grains/dairy products/meat/seafood/packaged foods. In addition, it covers image-based inspection systems, benchmark datasets and evaluation metrics along with key challenges faced which includes data scarcity, computational cost, interpretability and industrial deployment. In addition, intelligent food quality management by emerging technologies such as IoT, edge computing, blockchain and Explainable AI features.

  • Enhancing Dark Web Threat Monitoring With Transformer-Based Natural Language Processing

    by Manju Lata Joshi, Shikha Khullar, Mithlesh Arya

    The advent of decentralized and anonymized web infrastructures and the Dark Web specifically, has severely changed the situation with cyber crime. The Dark Web supports the black market through anonymity networks like Tor, I2P, and Freenet: ransomware-as-a-service, malware trading, financial fraud, and coordinated cyberattacks. These activities challenge cybersecurity agencies with serious concerns because traditional methods of key-based or even shallow NLP don't detect the contextual, multilingual, and adversarial language that is common in communications on the Dark Web. This chapter explores how Transformer-based Natural Language Processing (NLP) models overcome them. It examines the language peculiarities of Dark Web materials such as obfuscation, code-mixing, and language peculiarities of domain and emphasizes the limitations of the traditional methods of surveillance. The paper introduces an end-to-end threat intelligence pipeline based on the Transformer with the implementation of multiple stages such as Dark Web crawling, adversarial text preprocessing etc.

  • Understanding the Dark Web

    by Shikha Khullar, Manju Lata Joshi, Nitu Tank

    The Dark Web is an associative and developing digital ecosystem based on anonymity-defending technologies and cryptographic systems. In the current chapter, the author discusses the stratified structure of the Internet, as Surface Web, Deep Web, and Dark Web, as the anonymous communication systems and onion routing networks are being traced back to their historical evolution. It examines some of the most important anonymity systems, like Tor, I2P, and Freenet and considers the emergence of illegal markets and data markets and the cybercrime-as-a-service systems, data trafficking, and extremist networks. Local differences in applications and administration systems are sharply considered. Artificial intelligence in Dark Web policing has been given a special consideration, such as machine learning, blockchain analytics, network modeling, and large language models. Other emerging issues which are also pointed out in the chapter include adaptive criminal innovation, quantum-resistant encryption, and ethical governance of AI when making cross-border cyber investigations.

  • Ethical and Legal Implications in Medical AI

    by Shikha Khullar, Ibrar Ahmed, Sumit Kumar Kapoor, Manju Lata Joshi, Arun Arora

    The use of Large Language Models (LLMs) in the medical field, especially in clinical safety evaluation, has substantial prospective advantages to enhance the diagnostic performance, clinical decision-making process, and patient outcomes. Although these advantages exist, there are complicated ethical, legal, and governance issues with the introduction of AI in healthcare. This chapter critically reviews ethical implication of the use of LLM with special reference being given to clinical autonomy, algorithmic bias, informed consent, and accountability. It draws on real-life case studies to emphasize the role of biased training data and lack of transparency in decision-making in promoting health disparities and undermining trust between clinicians and patients. The analysis also outlines some of the crucial failure points in current legal and regulatory frameworks particularly on the aspect of the distribution of liability and privacy of the data that are ill suited to adapt to the adaptable and changing character of the AI systems.

  • Natural Language Processing With Hybrid Quantum Neural Networks

    by Manju Lata Joshi

    The recent developments in the field of deep learning have made Natural Language Processing (NLP) quite useful, but there are still issues of the complexity of computation, energy consumption, and modeling of complex semantics. Representations and optimization techniques that are quantum enhanced are available in quantum computing and provide new possibilities. Hybrid Quantum Neural Networks (HQNNs) combine quantum computational concepts and classical neural networks, allowing the introduction of a new era in intelligent language processing. The chapter provides the basics of quantum computing and neural NLP models and looks at the hybrid methods of semantic representation and learning. The text classification, sentiment analysis, named entity recognition, and machine translation tasks that are fundamental to NLP are examined through a hybrid quantum-neural approach. Such methods as quantum data encoding, quantum feature extraction and hybrid training approaches are described and the issues associated with scalability, noise, and hardware limitations are discussed.