English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Mapping speech tokens to the same feature space as text tokens has become theparadigm for the integration of speech modality into decoder-only largelanguage models (LLMs). An alternative approach is to use an encoder-decoderarchitecture that incorporates speech features through cross-attention. Thisapproach, however, has received less attention in the literature. In this work,we connect the Whisper encoder with ChatGLM3 and provide in-depth comparisonsof these two approaches using Chinese automatic speech recognition (ASR) andname entity recognition (NER) tasks. We evaluate them not only by conventionalmetrics like the F1 score but also by a novel fine-grained taxonomy of ASR-NERerrors. Our experiments reveal that encoder-decoder architecture outperformsdecoder-only architecture with a short context, while decoder-only architecturebenefits from a long context as it fully exploits all layers of the LLM. Byusing LLM, we significantly reduced the entity omission errors and improved theentity ASR accuracy compared to the Conformer baseline. Additionally, weobtained a state-of-the-art (SOTA) F1 score of 0.805 on the AISHELL-NER testset by using chain-of-thought (CoT) NER which first infers long-form ASRtranscriptions and then predicts NER labels.

Using Large Language Model for End-to-End Chinese ASR and NER