Toward Fine-Grained Speech Inpainting Forensics:A Dataset, Method, and Metric for Multi-Region Tampering Localization

• URL: http://arxiv.org/abs/2605.02223v1
• Date: Mon, 04 May 2026 04:54:29 GMT
• Title: Toward Fine-Grained Speech Inpainting Forensics:A Dataset, Method, and Metric for Multi-Region Tampering Localization
• Abstract summary: MIST (Multiregion Inpainting Speech Tampering) is a large-scale multilingual dataset spanning 6 languages.<n>ISA (Iterative Segment Analysis) is a backbone-agnostic framework that performs coarse-to-fine sliding-window classification.<n>SF1@tau is a segment-level F1 metric based on temporal IoU matching.
• Score: 7.260139045446439
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Recent advances in voice cloning and text-to-speech synthesis have made partial speech manipulation - where an adversary replaces a few words within an utterance to alter its meaning while preserving the speaker's identity - an increasingly realistic threat. Existing audio deepfake detection benchmarks focus on utterance-level binary classification or single-region tampering, leaving a critical gap in detecting and localizing multiple inpainted segments whose count is unknown a priori. We address this gap with three contributions. First, we introduce MIST (Multiregion Inpainting Speech Tampering), a large-scale multilingual dataset spanning 6 languages with 1-3 independently inpainted word-level segments per utterance, generated via LLM-guided semantic replacement and neural voice cloning, with fake content constituting only 2-7% of each utterance. Second, we propose ISA (Iterative Segment Analysis), a backbone-agnostic framework that performs coarse-to-fine sliding-window classification with gap-tolerant region proposal and boundary refinement to recover all tampered regions without prior knowledge of their count. Third, we define SF1@tau, a segment-level F1 metric based on temporal IoU matching that jointly evaluates region count accuracy and localization precision. Zero-shot evaluation reveals that partial inpainting at word granularity remains unsolved by existing deepfake detectors: utterance-level classifiers trained on fully synthesized speech assign near zero fake probability to MIST utterances where only 2-7% of content is manipulated. ISA consistently outperforms non-iterative baselines in this challenging setting, and the dataset, code, and evaluation toolkit are publicly released.

Related papers

• Split and Conquer Partial Deepfake Speech [1.8949616457421692]
  Partial deepfake speech detection requires identifying manipulated regions that may occur within short temporal portions of an otherwise bona fide utterance.<n>We propose a split-and-conquer framework that decomposes the problem into two stages: boundary detection and segment-level classification.<n>Experiments on the PartialSpoof benchmark demonstrate state-of-the-art performance across multiple temporal resolutions as well as at the utterance level.
  arXiv  Detail & Related papers  (2026-04-03T09:33:01Z)
• Adapting Self-Supervised Speech Representations for Cross-lingual Dysarthria Detection in Parkinson's Disease [72.0406069194794]
  Speech representations often encode language-dependent structure that can confound dysarthria detection.<n>We propose a representation-level language shift that aligns source-language self-supervised speech representations with the target-language distribution.<n>We evaluate the approach on oral DDK recordings from Parkinson's disease speech datasets in Czech, German, and Spanish.
  arXiv  Detail & Related papers  (2026-03-23T17:23:39Z)
• GLip: A Global-Local Integrated Progressive Framework for Robust Visual Speech Recognition [72.29071664964633]
  We propose GLip, a Global-Local Integrated Progressive framework designed for robust visual speech recognition (VSR)<n>GLip learns to align both global and local visual features with corresponding acoustic speech units using easily accessible audio-visual data.<n>In the second stage, we introduce a Contextual Enhancement Module (CEM) to dynamically integrate local features with relevant global context.
  arXiv  Detail & Related papers  (2025-09-19T14:36:01Z)
• LIRA: Inferring Segmentation in Large Multi-modal Models with Local Interleaved Region Assistance [54.683384204063934]
  Large multi-modal models (LMMs) struggle with inaccurate segmentation and hallucinated comprehension.<n>We propose LIRA, a framework that capitalizes on the complementary relationship between visual comprehension and segmentation.<n>LIRA achieves state-of-the-art performance in both segmentation and comprehension tasks.
  arXiv  Detail & Related papers  (2025-07-08T07:46:26Z)
• Advancing Topic Segmentation of Broadcasted Speech with Multilingual Semantic Embeddings [2.615008111842321]
  We introduce an end-to-end scheme for topic segmentation using semantic speech encoders. We propose a new benchmark for spoken news topic segmentation by utilizing a dataset featuring 1000 hours of publicly available recordings. Our results demonstrate that while the traditional pipeline approach achieves a state-of-the-art $P_k$ score of 0.2431 for English, our end-to-end model delivers a competitive $P_k$ score of 0.2564.
  arXiv  Detail & Related papers  (2024-09-10T05:24:36Z)
• Language-Universal Speech Attributes Modeling for Zero-Shot Multilingual Spoken Keyword Recognition [26.693942793501204]
  We propose a novel language-universal approach to end-to-end automatic spoken keyword recognition (SKR) Wav2Vec2.0 is used to generate robust speech representations, followed by a linear output layer to produce attribute sequences. A non-trainable pronunciation model then maps sequences of attributes into spoken keywords in a multilingual setting.
  arXiv  Detail & Related papers  (2024-06-04T16:59:11Z)
• SHiNe: Semantic Hierarchy Nexus for Open-vocabulary Object Detection [31.464227593768324]
  We introduce Semantic Hierarchy Nexus (SHiNe), a novel classifier that uses semantic knowledge from class hierarchies. SHiNe enhances robustness across diverse vocabulary granularities, achieving up to +31.9% mAP50 with ground truth hierarchies. SHiNe is training-free and can be seamlessly integrated with any off-the-shelf OvOD detector.
  arXiv  Detail & Related papers  (2024-05-16T12:42:06Z)
• CROP: Zero-shot Cross-lingual Named Entity Recognition with Multilingual Labeled Sequence Translation [113.99145386490639]
  Cross-lingual NER can transfer knowledge between languages via aligned cross-lingual representations or machine translation results. We propose a Cross-lingual Entity Projection framework (CROP) to enable zero-shot cross-lingual NER. We adopt a multilingual labeled sequence translation model to project the tagged sequence back to the target language and label the target raw sentence.
  arXiv  Detail & Related papers  (2022-10-13T13:32:36Z)
• On Cross-Lingual Retrieval with Multilingual Text Encoders [51.60862829942932]
  We study the suitability of state-of-the-art multilingual encoders for cross-lingual document and sentence retrieval tasks. We benchmark their performance in unsupervised ad-hoc sentence- and document-level CLIR experiments. We evaluate multilingual encoders fine-tuned in a supervised fashion (i.e., we learn to rank) on English relevance data in a series of zero-shot language and domain transfer CLIR experiments.
  arXiv  Detail & Related papers  (2021-12-21T08:10:27Z)
• Unsupervised Acoustic Unit Discovery by Leveraging a Language-Independent Subword Discriminative Feature Representation [31.87235700253597]
  This paper tackles automatically discovering phone-like acoustic units (AUD) from unlabeled speech data. We propose a two-stage approach: the first stage learns a subword-discriminative feature representation and the second stage applies clustering to the learned representation and obtains phone-like clusters as the discovered acoustic units.
  arXiv  Detail & Related papers  (2021-04-02T11:43:07Z)
• Video-based Sign Language Recognition without Temporal Segmentation [88.03159640595187]
  We propose a novel continuous sign recognition framework, which eliminates the preprocessing of temporal segmentation.<n>The proposed LS-HAN consists of three components: a two-stream Convolutional Neural Network (CNN) for video feature representation generation, a Latent Space for semantic gap bridging, and a Hierarchical Attention Network (HAN) for latent space based recognition.
  arXiv  Detail & Related papers  (2018-01-30T17:37:42Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.