Paper Citation Record · LEDGER

Model as Loss: A Self-Consistent Training Paradigm

As of 9 August 2026, this Paper Citation Record lists 36 of 36 outbound references and 1 inbound Pith citation observation for arXiv:2505.21156.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2505.21156 v1

Coverage vector

measured 36 of 36 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-07T13:42:42.235277Z

measured 37 of 37 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-09T06:31:02.800959+00:00

measured 1 of 1 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-07T13:42:39.417023Z

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: pith, observed 2026-08-07T13:42:43.278084Z

Reference resolution

36 of 36 outbound references displayed

verified exact0
verified fuzzy21
unresolved12
parse uncertain0
malformed identifier0
metadata mismatch3

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 0e687cf0-439d-4192-b42c-ccf3fb14be10 · outbound

This paper cites A critical component in training enhancement models is the choice of loss function, which directly influences the quality and generaliza- tion of enhanced output [6, 7].

Model as Loss: A Self-Consistent Training Paradigm A critical component in training enhancement models is the choice of loss function, which directly influences the quality and generaliza- tion of enhanced output [6, 7]

Reference 1

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:46.878072Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 150fbb1c-0708-405f-8fed-afe245e4acbc · outbound

This paper cites Model as Loss: A Self-Consistent Training Paradigm.

Model as Loss: A Self-Consistent Training Paradigm Model as Loss: A Self-Consistent Training Paradigm

Reference 2

Resolution: metadata mismatch; local_arxiv, observed 2026-08-07T13:42:43.386551Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation bd4daa8e-3902-4d8a-8f4d-024057d3f9e2 · outbound

This paper cites It has an encoder that extracts relevant fea- tures and passes them into a first-stage decoder.

Model as Loss: A Self-Consistent Training Paradigm It has an encoder that extracts relevant fea- tures and passes them into a first-stage decoder

Reference 3

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:46.746243Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation b2057606-70ce-4716-bf5e-5c4417b66d81 · outbound

This paper cites Oursmal−dynamic achieves the best performance across all NISQA metrics, while Ours mal−frozen leads in all intrusive metrics.

Model as Loss: A Self-Consistent Training Paradigm Oursmal−dynamic achieves the best performance across all NISQA metrics, while Ours mal−frozen leads in all intrusive metrics

Reference 4

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:46.577644Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation adc5ea9c-781c-4e2c-8526-d1ed84ddb60b · outbound

This paper cites By align- ing the loss with the model’s task-specific feature space,MAL overcomes the limitations of traditional handcrafted and pre- trained deep feature losses.

Model as Loss: A Self-Consistent Training Paradigm By align- ing the loss with the model’s task-specific feature space,MAL overcomes the limitations of traditional handcrafted and pre- trained deep feature losses

Reference 5

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:46.409394Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation da2cb6ea-fbfe-4965-8815-eb2b6cd21079 · outbound

This paper cites We also thank Sai Dhawal Phaye for discussions during the early stages of MAL, and Kanav Sabharwal for his feedback on the writing.

Model as Loss: A Self-Consistent Training Paradigm We also thank Sai Dhawal Phaye for discussions during the early stages of MAL, and Kanav Sabharwal for his feedback on the writing

Reference 6

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:46.262251Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation db361ab7-6d80-4eb1-8980-21fff662b73f · outbound

This paper cites Benesty, S.

Model as Loss: A Self-Consistent Training Paradigm Benesty, S

Reference 7

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:46.099919Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 5a19d199-cecf-4d1e-99a9-df58fa5eaf18 · outbound

This paper cites The INTERSPEECH 2020 Deep Noise Suppression Challenge: Datasets, Subjective Testing Framework, and Challenge Results.

Model as Loss: A Self-Consistent Training Paradigm The INTERSPEECH 2020 Deep Noise Suppression Challenge: Datasets, Subjective Testing Framework, and Challenge Results

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:39.659525Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 59f19fad-0952-4a46-97a4-a0f158c95a2b · outbound

This paper cites FINALLY: fast and universal speech enhancement with studio- like quality,.

Model as Loss: A Self-Consistent Training Paradigm FINALLY: fast and universal speech enhancement with studio- like quality,

Reference 9

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:45.961420Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 91f9ca76-1716-46ee-bd05-0ca97149b6b8 · outbound

This paper cites Icassp 2023 deep noise suppression challenge,.

Model as Loss: A Self-Consistent Training Paradigm Icassp 2023 deep noise suppression challenge,

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:45.814651Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 0cfe1aad-1d20-49ea-99d6-fa17a36c6706 · outbound

This paper cites Wavlm: Large-scale self-supervised pre-training for full stack speech processing,.

Model as Loss: A Self-Consistent Training Paradigm Wavlm: Large-scale self-supervised pre-training for full stack speech processing,

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:40.099883Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c54b29f9-6209-4705-8ce6-54c93371eb01 · outbound

This paper cites DeepFilterNet2: Towards real-time speech enhancement on em- bedded devices for full-band audio,.

Model as Loss: A Self-Consistent Training Paradigm DeepFilterNet2: Towards real-time speech enhancement on em- bedded devices for full-band audio,

Reference 12

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:45.637746Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation cd0e01e3-40a0-46ae-bca0-6769c5d08127 · outbound

This paper cites Speech denoising with deep feature losses,.

Model as Loss: A Self-Consistent Training Paradigm Speech denoising with deep feature losses,

Reference 13

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:45.457857Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation edc8d6a3-73ac-4ff2-b995-3ade4958300a · outbound

This paper cites A consolidated view of loss functions for supervised deep learning-based speech enhancement,.

Model as Loss: A Self-Consistent Training Paradigm A consolidated view of loss functions for supervised deep learning-based speech enhancement,

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:39.901561Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 42003d0f-b1e2-4356-abe8-327c4ef1f3d6 · outbound

This paper cites auraloss: Audio focused loss functions in PyTorch,.

Model as Loss: A Self-Consistent Training Paradigm auraloss: Audio focused loss functions in PyTorch,

Reference 15

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:45.308332Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation ebdb0d5a-8ba9-4425-9e52-04e3060ee408 · outbound

This paper cites A deep learning loss function based on the perceptual evaluation of the speech quality,.

Model as Loss: A Self-Consistent Training Paradigm A deep learning loss function based on the perceptual evaluation of the speech quality,

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:45.148533Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation d7db6738-659c-470b-b7db-e16ae999cec3 · outbound

This paper cites On loss func- tions for supervised monaural time-domain speech enhancement,.

Model as Loss: A Self-Consistent Training Paradigm On loss func- tions for supervised monaural time-domain speech enhancement,

Reference 17

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:44.974447Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation fcc36071-f261-4f1b-bea1-bbb06ebcea90 · outbound

This paper cites Nisqa: A deep cnn-self-attention model for multidimensional speech quality prediction with crowdsourced datasets,.

Model as Loss: A Self-Consistent Training Paradigm Nisqa: A deep cnn-self-attention model for multidimensional speech quality prediction with crowdsourced datasets,

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:41.253378Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6c47dc4f-ccf0-487f-9318-a04e7d3f7167 · outbound

This paper cites wav2vec: Unsupervised Pre-training for Speech Recognition.

Model as Loss: A Self-Consistent Training Paradigm wav2vec: Unsupervised Pre-training for Speech Recognition

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:40.142844Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 303cc871-b196-4e99-a144-c27aab83368c · outbound

This paper cites wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations.

Model as Loss: A Self-Consistent Training Paradigm wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:40.191267Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 44abafa9-faad-436e-97c9-c881514f32e3 · outbound

This paper cites Speechlmscore: Evaluating speech generation using speech language model,.

Model as Loss: A Self-Consistent Training Paradigm Speechlmscore: Evaluating speech generation using speech language model,

Reference 21

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:44.788952Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation dd060110-7b75-4eb9-adb8-ec375a6cfb94 · outbound

This paper cites Enhancing lora reception with generative models: Channel-aware denoising of loraphy signals,.

Model as Loss: A Self-Consistent Training Paradigm Enhancing lora reception with generative models: Channel-aware denoising of loraphy signals,

Reference 22

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:44.314166Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 4464a80b-d1d0-4913-b132-de41569bc7ce · outbound

This paper cites Icassp 2022 deep noise suppression challenge,.

Model as Loss: A Self-Consistent Training Paradigm Icassp 2022 deep noise suppression challenge,

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:40.490769Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation be6f4b9e-2860-4bbe-ae6c-4c34b7946efc · outbound

This paper cites ICASSP 2022 Deep Noise Suppression Challenge.

Model as Loss: A Self-Consistent Training Paradigm ICASSP 2022 Deep Noise Suppression Challenge

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:40.643307Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5bc3bf02-a5de-4c8d-adfa-f947ccf833b0 · outbound

This paper cites URGENT Challenge: Universality, Robustness, and Generalizability For Speech Enhancement.

Model as Loss: A Self-Consistent Training Paradigm URGENT Challenge: Universality, Robustness, and Generalizability For Speech Enhancement

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:40.758830Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 54651185-5f5a-464d-9925-227b7caeefb0 · outbound

This paper cites Icassp 2024 speech signal improvement challenge,.

Model as Loss: A Self-Consistent Training Paradigm Icassp 2024 speech signal improvement challenge,

Reference 26

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:44.658530Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 4f43fbe4-1f5f-4c6c-becd-d29978df1be6 · outbound

This paper cites SCOREQ: Speech Quality Assessment with Contrastive Regression.

Model as Loss: A Self-Consistent Training Paradigm SCOREQ: Speech Quality Assessment with Contrastive Regression

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:41.419789Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6b0ccf30-8112-4dd6-81d9-42553d8ac460 · outbound

This paper cites Do- main adaptation and autoencoder-based unsupervised speech enhancement,.

Model as Loss: A Self-Consistent Training Paradigm Do- main adaptation and autoencoder-based unsupervised speech enhancement,

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:41.587230Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f11b3ce9-39c7-4833-b654-1f33f4406337 · outbound

This paper cites Medical image denoising using convolutional de- noising autoencoders,.

Model as Loss: A Self-Consistent Training Paradigm Medical image denoising using convolutional de- noising autoencoders,

Reference 31

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:44.499649Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 9543ab20-8336-4368-9896-72f919dc0d86 · outbound

This paper cites Neural cascade architecture for multi- channel acoustic echo suppression,.

Model as Loss: A Self-Consistent Training Paradigm Neural cascade architecture for multi- channel acoustic echo suppression,

Reference 33

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:44.162551Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation ecab8680-b463-473e-a1b2-1bb8089e6e9e · outbound

This paper cites Task splitting for dnn-based acous- tic echo and noise removal,.

Model as Loss: A Self-Consistent Training Paradigm Task splitting for dnn-based acous- tic echo and noise removal,

Reference 34

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:43.979802Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 48154d5e-cad9-406d-b0e0-735c677de124 · outbound

This paper cites Deep learning on image denoising: An overview,.

Model as Loss: A Self-Consistent Training Paradigm Deep learning on image denoising: An overview,

Reference 35

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:43.771611Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 87513493-1128-4794-813d-244aeee0a603 · outbound

This paper cites A review of the deep learning methods for medical images super resolution problems,.

Model as Loss: A Self-Consistent Training Paradigm A review of the deep learning methods for medical images super resolution problems,

Reference 36

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T13:42:43.580443Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation 908ec97b-432e-4761-956d-d96b12707b8a · outbound

This paper cites SpeechLMScore: Evaluating speech generation using speech language model.

Model as Loss: A Self-Consistent Training Paradigm SpeechLMScore: Evaluating speech generation using speech language model

Reference 2022

Resolution: metadata mismatch; local_arxiv, observed 2026-08-07T13:42:42.972593Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Observation bd69fe6a-349b-4f5f-be5e-4ca7af14ca70 · outbound

This paper cites ICASSP 2023 Deep Noise Suppression Challenge.

Model as Loss: A Self-Consistent Training Paradigm ICASSP 2023 Deep Noise Suppression Challenge

Reference 2023

Resolution: unresolved; no resolver link, observed 2026-08-07T13:42:39.748666Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3646aa66-9ea5-4e07-8b1c-6ca5d0b6f0e0 · outbound

This paper cites ICASSP 2024 Speech Signal Improvement Challenge.

Model as Loss: A Self-Consistent Training Paradigm ICASSP 2024 Speech Signal Improvement Challenge

Reference 2024

Resolution: metadata mismatch; local_arxiv, observed 2026-08-07T13:42:42.718972Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.

Pith citing papers

Observation 150fbb1c-0708-405f-8fed-afe245e4acbc · inbound