Diptesh Kanojia

Bridging Domains for Automatic Post-Editing: A Classifier-Guided Multi-Domain Adaptation Framework

Quality-Aware Self-Correcting Speech Translation on an Edge Device

We present a fully offline speech-to-speech translation pipeline that runs on a Jetson Nano (4 GB) and corrects its own weak translations without retraining. A Whisper-tiny ASR feeds an Opus-MT translator; multilingual BERT cosine similarity acts as a Quality Estimation (QE) gate, triggering a secondary-pass correction...

Beyond Relevance: Structured Semantic Supervision for Product Search with LLM-Augmented Annotations

Blations show that most of the oracle improvement comes from post-edited explanations and annotator comments rather than the scalar centrality feature, suggesting that LLMs are most useful for exposing and approximating structured semantic supervision rather than replacing human judgment directly.

INDICQE-APE: A Benchmark for Quality Estimation and Automatic Post-Editing for Indic Languages

The WMT 2020–2024 shared-task lineage with an extended English–Malayalam resource is consolidated into INDICQE-APE, with up to four label types aligned on the same segment, a direct assessment, a human post-edit, word-level OK/BAD tags and an error explanation, and a test set stratified over four difficulty axes.

Decodable but Misrouted: Sparse Features Uncover a Readout Gap in Vision-Language Models for Harmful Meme Detection

Robustness controls show that the signal is not explained solely by accompanying OCR and depends on paired visual evidence, and that it extends beyond English.

In the Blind: Building Pseudo-References for MT Evaluation

The WMT26 General MT task evaluates systems on 10 language pairs that have no human references (neither translated from scratch nor post-edited from MT output by humans) and how the pseudo-references were built is described.

DiaLLM: An Investigation into the Robustness-Generation Gap in English Dialect Adaptation

DiaLLM is introduced, which continually pretrains three open-weight language model families on the International Corpus of English and applies implicit and explicit post-training paradigms, each combined with three model alignment strategies, giving the first controlled comparison of these components across Australian,...

IndicQE-APE: A Consolidated Benchmark for Quality Estimation and Automatic Post-Editing for Indic Languages

The WMT 2020-2024 shared-task lineage with an extended English-Malayalam resource is consolidated into IndicQE-APE, with up to four label types aligned on the same segment, a direct assessment, a human post-edit, word-level tags and an error explanation, and a test set stratified over four difficulty axes.

SyncDreamer : Controllable and Expressive Avatar Generation Beyond the Talking Head

SyncDreamer is presented, a unified diffusion Transformer framework that generates identity-preserving and emotionally expressive talking avatars from only a single image, speech audio, and text prompt, and an RL-based Cross-Modal Prompt Enhancer grounding textual cues in visual context for fine-grained motion con-trol...