Takafumi Koshinaka

Kosuke Takemoto, Takafumi Koshinaka

Diffusion-based virtual try-on methods achieve photorealistic synthesis through cross-attention mechanisms that transfer garment features to target body regions. However, these approaches rely on implicit learning of spatial correspondences, struggling to preserve fine details such as text and illustrations. We propose a novel approach, which we call SIFT-VTON, that utilizes SIFT keypoint matching to provide explicit geometric guidance for diffusion-based virtual try-on. Our method applies domain-specific filtering to SIFT keypoint matches between garment and person images, then converts these correspondences into spatial probability distributions that supervise cross-attention layers during training. This explicit supervision guides the model to learn precise spatial alignment, concentrating attention on geometrically consistent garment regions. Experiments on the VITON-HD dataset demonstrate significant improvements on unpaired metrics while maintaining competitive paired reconstruction metrics. Qualitative comparisons show superior preservation of text clarity and pattern alignment. Attention visualizations confirm that our method produces sharply focused attention on relevant garment details. This work demonstrates that classical geometric correspondence methods can effectively enhance modern diffusion models for conditional synthesis tasks. The source code will be available at https://github.com/takesukeDS/SIFT-VTON.

Kosuke Takemoto, Takafumi Koshinaka

Virtual try-on systems have significant potential in e-commerce, allowing customers to visualize garments on themselves. Existing image-based methods fall into two categories: those that directly warp garment-images onto person-images (explicit warping), and those using cross-attention to reconstruct given garments (implicit warping). Explicit warping preserves garment details but often produces unrealistic output, while implicit warping achieves natural reconstruction but struggles with fine details. We propose HYB-VITON, a novel approach that combines the advantages of each method and includes both a preprocessing pipeline for warped garments and a novel training option. These components allow us to utilize beneficial regions of explicitly warped garments while leveraging the natural reconstruction of implicit warping. A series of experiments demonstrates that HYB-VITON preserves garment details more faithfully than recent diffusion-based methods, while producing more realistic results than a state-of-the-art explicit warping method.

Honori Udo, Takafumi Koshinaka

We revisit language bottleneck models as an approach to ensuring the explainability of deep learning models for image classification. Because of inevitable information loss incurred in the step of converting images into language, the accuracy of language bottleneck models is considered to be inferior to that of standard black-box models. Recent image captioners based on large-scale foundation models of Vision and Language, however, have the ability to accurately describe images in verbal detail to a degree that was previously believed to not be realistically possible. In a task of disaster image classification, we experimentally show that a language bottleneck model that combines a modern image captioner with a pre-trained language model can achieve image classification accuracy that exceeds that of black-box models. We also demonstrate that a language bottleneck model and a black-box model may be thought to extract different features from images and that fusing the two can create a synergistic effect, resulting in even higher classification accuracy.

Honori Udo, Takafumi Koshinaka

Image captioning, a.k.a. "image-to-text," which generates descriptive text from given images, has been rapidly developing throughout the era of deep learning. To what extent is the information in the original image preserved in the descriptive text generated by an image captioner? To answer that question, we have performed experiments involving the classification of images from descriptive text alone, without referring to the images at all, and compared results with those from standard image-based classifiers. We have evaluate several image captioning models with respect to a disaster image classification task, CrisisNLP, and show that descriptive text classifiers can sometimes achieve higher accuracy than standard image-based classifiers. Further, we show that fusing an image-based classifier with a descriptive text classifier can provide improvement in accuracy.

Qiongqiong Wang, Kong Aik Lee, Takafumi Koshinaka et al.

This paper proposes the use of two task-aware warping factors in mask-based speech enhancement (SE). One controls the balance between speech-maintenance and noise-removal in training phases, while the other controls SE power applied to specific downstream tasks in testing phases. Our intention is to alleviate the problem that SE systems trained to improve speech quality often fail to improve other downstream tasks, such as automatic speaker verification (ASV) and automatic speech recognition (ASR), because they do not share the same objects. It is easy to apply the proposed dual-warping factors approach to any mask-based SE method, and it allows a single SE system to handle multiple tasks without task-dependent training. The effectiveness of our proposed approach has been confirmed on the SITW dataset for ASV evaluation and the LibriSpeech dataset for ASR and speech quality evaluations of 0-20dB. We show that different warping values are necessary for a single SE to achieve optimal performance w.r.t. the three tasks. With the use of task-dependent warping factors, speech quality was improved by an 84.7% PESQ increase, ASV had a 22.4% EER reduction, and ASR had a 52.2% WER reduction, on 0dB speech. The effectiveness of the task-dependent warping factors were also cross-validated on VoxCeleb-1 test set for ASV and LibriSpeech dev-clean set for ASV and quality evaluations. The proposed method is highly effective and easy to apply in practice.

Kong Aik Lee, Qiongqiong Wang, Takafumi Koshinaka

We present a Bayesian formulation for deep speaker embedding, wherein the xi-vector is the Bayesian counterpart of the x-vector, taking into account the uncertainty estimate. On the technology front, we offer a simple and straightforward extension to the now widely used x-vector. It consists of an auxiliary neural net predicting the frame-wise uncertainty of the input sequence. We show that the proposed extension leads to substantial improvement across all operating points, with a significant reduction in error rates and detection cost. On the theoretical front, our proposal integrates the Bayesian formulation of linear Gaussian model to speaker-embedding neural networks via the pooling layer. In one sense, our proposal integrates the Bayesian formulation of the i-vector to that of the x-vector. Hence, we refer to the embedding as the xi-vector, which is pronounced as /zai/ vector. Experimental results on the SITW evaluation set show a consistent improvement of over 17.5% in equal-error-rate and 10.9% in minimum detection cost.

Qiongqiong Wang, Kong Aik Lee, Takafumi Koshinaka

This paper presents a simple but effective method that uses multi-resolution feature maps with convolutional neural networks (CNNs) for anti-spoofing in automatic speaker verification (ASV). The central idea is to alleviate the problem that the feature maps commonly used in anti-spoofing networks are insufficient for building discriminative representations of audio segments, as they are often extracted by a single-length sliding window. Resulting trade-offs between time and frequency resolutions restrict the information in single spectrograms. The proposed method improves both frequency resolution and time resolution by stacking multiple spectrograms that are extracted using different window lengths. These are fed into a convolutional neural network in the form of multiple channels, making it possible to extract more information from input signals while only marginally increasing computational costs. The efficiency of the proposed method has been conformed on the ASVspoof 2019 database. We show that the use of the proposed multiresolution inputs consistently outperforms that of score fusion across different CNN architectures. Moreover, computational cost remains small.

Qiongqiong Wang, Koji Okabe, Kong Aik Lee et al.

This paper proposes a generalized framework for domain adaptation of Probabilistic Linear Discriminant Analysis (PLDA) in speaker recognition. It not only includes several existing supervised and unsupervised domain adaptation methods but also makes possible more flexible usage of available data in different domains. In particular, we introduce here the two new techniques described below. (1) Correlation-alignment-based interpolation and (2) covariance regularization. The proposed correlation-alignment-based interpolation method decreases minCprimary up to 30.5% as compared with that from an out-of-domain PLDA model before adaptation, and minCprimary is also 5.5% lower than with a conventional linear interpolation method with optimal interpolation weights. Further, the proposed regularization technique ensures robustness in interpolations w.r.t. varying interpolation weights, which in practice is essential.

Ville Vestman, Kong Aik Lee, Tomi H. Kinnunen et al.

Speaker embeddings are continuous-value vector representations that allow easy comparison between voices of speakers with simple geometric operations. Among others, i-vector and x-vector have emerged as the mainstream methods for speaker embedding. In this paper, we illustrate the use of modern computation platform to harness the benefit of GPU acceleration for i-vector extraction. In particular, we achieve an acceleration of 3000 times in frame posterior computation compared to real time and 25 times in training the i-vector extractor compared to the CPU baseline from Kaldi toolkit. This significant speed-up allows the exploration of ideas that were hitherto impossible. In particular, we show that it is beneficial to update the universal background model (UBM) and re-compute frame alignments while training the i-vector extractor. Additionally, we are able to study different variations of i-vector extractors more rigorously than before. In this process, we reveal some undocumented details of Kaldi's i-vector extractor and show that it outperforms the standard formulation by a margin of 1 to 2% when tested with VoxCeleb speaker verification protocol. All of our findings are asserted by ensemble averaging the results from multiple runs with random start.

Kong Aik Lee, Ville Hautamaki, Tomi Kinnunen et al.

The I4U consortium was established to facilitate a joint entry to NIST speaker recognition evaluations (SRE). The latest edition of such joint submission was in SRE 2018, in which the I4U submission was among the best-performing systems. SRE'18 also marks the 10-year anniversary of I4U consortium into NIST SRE series of evaluation. The primary objective of the current paper is to summarize the results and lessons learned based on the twelve sub-systems and their fusion submitted to SRE'18. It is also our intention to present a shared view on the advancements, progresses, and major paradigm shifts that we have witnessed as an SRE participant in the past decade from SRE'08 to SRE'18. In this regard, we have seen, among others, a paradigm shift from supervector representation to deep speaker embedding, and a switch of research challenge from channel compensation to domain adaptation.

Kong Aik Lee, Qiongqiong Wang, Takafumi Koshinaka

State-of-the-art speaker recognition systems comprise an x-vector (or i-vector) speaker embedding front-end followed by a probabilistic linear discriminant analysis (PLDA) backend. The effectiveness of these components relies on the availability of a large collection of labeled training data. In practice, it is common that the domains (e.g., language, demographic) in which the system are deployed differs from that we trained the system. To close the gap due to the domain mismatch, we propose an unsupervised PLDA adaptation algorithm to learn from a small amount of unlabeled in-domain data. The proposed method was inspired by a prior work on feature-based domain adaptation technique known as the correlation alignment (CORAL). We refer to the model-based adaptation technique proposed in this paper as CORAL+. The efficacy of the proposed technique is experimentally validated on the recent NIST 2016 and 2018 Speaker Recognition Evaluation (SRE'16, SRE'18) datasets.

Qiongqiong Wang, Koji Okabe, Kong Aik Lee et al.

This paper presents an experimental study on deep speaker embedding with an attention mechanism that has been found to be a powerful representation learning technique in speaker recognition. In this framework, an attention model works as a frame selector that computes an attention weight for each frame-level feature vector, in accord with which an utterancelevel representation is produced at the pooling layer in a speaker embedding network. In general, an attention model is trained together with the speaker embedding network on a single objective function, and thus those two components are tightly bound to one another. In this paper, we consider the possibility that the attention model might be decoupled from its parent network and assist other speaker embedding networks and even conventional i-vector extractors. This possibility is demonstrated through a series of experiments on a NIST Speaker Recognition Evaluation (SRE) task, with 9.0% EER reduction and 3.8% min_Cprimary reduction when the attention weights are applied to i-vector extraction. Another experiment shows that DNN-based soft voice activity detection (VAD) can be effectively combined with the attention mechanism to yield further reduction of minCprimary by 6.6% and 1.6% in deep speaker embedding and i-vector systems, respectively.

Koji Okabe, Takafumi Koshinaka, Koichi Shinoda

This paper proposes attentive statistics pooling for deep speaker embedding in text-independent speaker verification. In conventional speaker embedding, frame-level features are averaged over all the frames of a single utterance to form an utterance-level feature. Our method utilizes an attention mechanism to give different weights to different frames and generates not only weighted means but also weighted standard deviations. In this way, it can capture long-term variations in speaker characteristics more effectively. An evaluation on the NIST SRE 2012 and the VoxCeleb data sets shows that it reduces equal error rates (EERs) from the conventional method by 7.5% and 8.1%, respectively.