Publications

2026 · Elsevier

Cross-Lingual Semantic Alignment and Multi-Objective Reranking for Multilingual Abstractive Summarization

Parshav Singla, Jasmeet Singh, Gurpal Singh Chhabra

Knowledge-Based Systems. Vol. 351, article 116785. Elsevier, 2026.

Q1 journal · Impact Factor 8.0 · CiteScore 13.7

Multilingual summarisers often write fluent summaries that drift from what the article says. We fine-tune mT5-Large on XL-Sum in five languages with an alignment loss that ties each article to its translation, then rerank candidate summaries by factual entailment (mDeBERTa-v3) and semantic centrality (MPNet).

• Cross-lingual semantic consistency improved for every language pair, most strongly for English and Swahili (+0.16).
• Expert human ratings rose from 3.42 to 4.13 out of 5.
• Languages: English, French, Spanish, Turkish and Swahili.

• Multilingual NLP
• Summarisation
• Factual consistency
• mT5

Cite: Singla, P., Singh, J., & Chhabra, G. S. (2026). Cross-Lingual Semantic Alignment and Multi-Objective Reranking for Multilingual Abstractive Summarization. Knowledge-Based Systems, vol. 351, article 116785, 2026. https://doi.org/10.1016/j.knosys.2026.116785

2025 · Springer, Singapore

Bridging the Gap: Converting Read Text to Conversational Dialogue

Parshav Singla, Agnik Banerjee, Aaditya Arora, Shruti Aggarwal, Anil Kumar Verma, C. M. Vikram, Raj Prakash Gohil, Gopal Kumar Agarwal

Innovative Computing and Communications (ICICC 2025). Lecture Notes in Networks and Systems, vol. 1431, pp. 543–556. Springer, Singapore, 2025.

Read speech sounds flat in a dialogue. PACC adjusts the intonation, stress and rhythm of read speech with a deep neural network and resynthesises it with a HiFi-GAN vocoder. The work came out of the Samsung PRISM programme, where we recorded 100 read and 100 conversational speech samples for training.

• Mean opinion score rose to 4.2 against a 3.6 baseline.
• First-author paper, presented at ICICC 2025 in New Delhi.

• Speech synthesis
• Prosody
• HiFi-GAN
• Conversational speech

Cite: Singla, P., Banerjee, A., Arora, A., Aggarwal, S., Verma, A. K., Vikram, C. M., Gohil, R. P., & Agarwal, G. K. (2025). Bridging the Gap: Converting Read Text to Conversational Dialogue. Innovative Computing and Communications (ICICC 2025), Lecture Notes in Networks and Systems, vol. 1431, pp. 543–556. Springer, Singapore, 2025. https://doi.org/10.1007/978-981-96-6681-2_38