The objective of neural network (NN) robustness certification is to determine if a NN changes its predictions when mutations are made to its inputs. While most certification research studies pixel-level or a few geometrical-level and blurring operations over images, this paper proposes a novel framework, GCERT, which certifies NN robustness under a precise and unified form of diverse semantic-level image mutations. We formulate a comprehensive set of semantic-level image mutations uniformly as certain directions in the latent space of generative models. We identify two key properties, independence and continuity, that convert the latent space into a precise and analysis-friendly input space representation for certification. GCERT can be smoothly integrated with de facto complete, incomplete, or quantitative certification frameworks. With its precise input space representation, GCERT enables for the first time complete NN robustness certification with moderate cost under diverse semantic-level input mutations, such as weather-filter, style transfer, and perceptual changes (e.g., opening/closing eyes). We show that GCERT enables certifying NN robustness under various common and security-sensitive scenarios like autonomous driving.
Automatic segmentation of fluid in OCT (Optical Coherence Tomography) images is beneficial for ophthalmologists to make an accurate diagnosis. Currently, data-driven convolutional neural networks (CNNs) have achieved great success in OCT fluid segmentation. However, obtaining pixel-level masks of OCT images is time-consuming and requires expertise. The popular weakly-supervised strategy is to generate noisy pseudo-labels from weak annotations, but the noise information introduced may mislead the model training. To address this issue, (i) we propose a superpixel-guided method for generating noisy labels from weak point annotations, called Point to Noisy by Superpixel (PNS), which limits the network from over-fitting noise by assigning low confidence to suspiciously noisy label pixels, and (ii) we propose a Two-Stage Mean-Teacher-assisted Confident Learning (2SMTCL) method based on MTCL for multi-category OCT fluid segmentation, which alleviates the uncertainty and computing power consumption introduced by the real-time characterization noise of MTCL. For evaluation, we have constructed a 2D OCT fluid segmentation dataset. Compared with other state-of-art label-denoising methods, comprehensive experimental results demonstrate that the proposed method can achieve excellent performance in OCT fluid segmentation as well as label denoising. Our study provides an efficient, accurate, and practical solution for fluid segmentation of OCT images, which is expected to have a positive impact on the diagnosis and treatment of patients in the field of ophthalmology.
Autonomous parking (AP) is an emering technique to navigate an intelligent vehicle to a parking space without any human intervention. Existing AP methods based on mathematical optimization or machine learning may lead to potential failures due to either excessive execution time or lack of generalization. To fill this gap, this paper proposes an integrated constrained optimization and imitation learning (iCOIL) approach to achieve efficient and reliable AP. The iCOIL method has two candidate working modes, i.e., CO and IL, and adopts a hybrid scenario analysis (HSA) model to determine the better mode under various scenarios. We implement and verify iCOIL on the Macao Car Racing Metaverse (MoCAM) platform. Results show that iCOIL properly adapts to different scenarios during the entire AP procedure, and achieves significantly larger success rates than other benchmarks.
Training a Named Entity Recognition (NER) model often involves fixing a taxonomy of entity types. However, requirements evolve and we might need the NER model to recognize additional entity types. A simple approach is to re-annotate entire dataset with both existing and additional entity types and then train the model on the re-annotated dataset. However, this is an extremely laborious task. To remedy this, we propose a novel approach called Partial Label Model (PLM) that uses only partially annotated datasets. We experiment with 6 diverse datasets and show that PLM consistently performs better than most other approaches (0.5 - 2.5 F1), including in novel settings for taxonomy expansion not considered in prior work. The gap between PLM and all other approaches is especially large in settings where there is limited data available for the additional entity types (as much as 11 F1), thus suggesting a more cost effective approaches to taxonomy expansion.
There has been an increasing interest in enhancing the fairness of machine learning (ML). Despite the growing number of fairness-improving methods, we lack a systematic understanding of the trade-offs among factors considered in the ML pipeline when fairness-improving methods are applied. This understanding is essential for developers to make informed decisions regarding the provision of fair ML services. Nonetheless, it is extremely difficult to analyze the trade-offs when there are multiple fairness parameters and other crucial metrics involved, coupled, and even in conflict with one another. This paper uses causality analysis as a principled method for analyzing trade-offs between fairness parameters and other crucial metrics in ML pipelines. To ractically and effectively conduct causality analysis, we propose a set of domain-specific optimizations to facilitate accurate causal discovery and a unified, novel interface for trade-off analysis based on well-established causal inference methods. We conduct a comprehensive empirical study using three real-world datasets on a collection of widelyused fairness-improving techniques. Our study obtains actionable suggestions for users and developers of fair ML. We further demonstrate the versatile usage of our approach in selecting the optimal fairness-improving method, paving the way for more ethical and socially responsible AI technologies.
Voice conversion is an increasingly popular technology, and the growing number of real-time applications requires models with streaming conversion capabilities. Unlike typical (non-streaming) voice conversion, which can leverage the entire utterance as full context, streaming voice conversion faces significant challenges due to the missing future information, resulting in degraded intelligibility, speaker similarity, and sound quality. To address this challenge, we propose DualVC, a dual-mode neural voice conversion approach that supports both streaming and non-streaming modes using jointly trained separate network parameters. Furthermore, we propose intra-model knowledge distillation and hybrid predictive coding (HPC) to enhance the performance of streaming conversion. Additionally, we incorporate data augmentation to train a noise-robust autoregressive decoder, improving the model's performance on long-form speech conversion. Experimental results demonstrate that the proposed model outperforms the baseline models in the context of streaming voice conversion, while maintaining comparable performance to the non-streaming topline system that leverages the complete context, albeit with a latency of only 252.8 ms.
We explore how weak supervision on abundant unlabeled data can be leveraged to improve few-shot performance in aspect-based sentiment analysis (ABSA) tasks. We propose a pipeline approach to construct a noisy ABSA dataset, and we use it to adapt a pre-trained sequence-to-sequence model to the ABSA tasks. We test the resulting model on three widely used ABSA datasets, before and after fine-tuning. Our proposed method preserves the full fine-tuning performance while showing significant improvements (15.84% absolute F1) in the few-shot learning scenario for the harder tasks. In zero-shot (i.e., without fine-tuning), our method outperforms the previous state of the art on the aspect extraction sentiment classification (AESC) task and is, additionally, capable of performing the harder aspect sentiment triplet extraction (ASTE) task.
This paper proposes a novel Attention-based Encoder-Decoder network for End-to-End Neural speaker Diarization (AED-EEND). In AED-EEND system, we incorporate the target speaker enrollment information used in target speaker voice activity detection (TS-VAD) to calculate the attractor, which can mitigate the speaker permutation problem and facilitate easier model convergence. In the training process, we propose a teacher-forcing strategy to obtain the enrollment information using the ground-truth label. Furthermore, we propose three heuristic decoding methods to identify the enrollment area for each speaker during the evaluation process. Additionally, we enhance the attractor calculation network LSTM used in the end-to-end encoder-decoder based attractor calculation (EEND-EDA) system by incorporating an attention-based model. By utilizing such an attention-based attractor decoder, our proposed AED-EEND system outperforms both the EEND-EDA and TS-VAD systems with only 0.5s of enrollment data.
EXplainable AI (XAI) is an essential topic to improve human understanding of deep neural networks (DNNs) given their black-box internals. For computer vision tasks, mainstream pixel-based XAI methods explain DNN decisions by identifying important pixels, and emerging concept-based XAI explore forming explanations with concepts (e.g., a head in an image). However, pixels are generally hard to interpret and sensitive to the imprecision of XAI methods, whereas "concepts" in prior works require human annotation or are limited to pre-defined concept sets. On the other hand, driven by large-scale pre-training, Segment Anything Model (SAM) has been demonstrated as a powerful and promotable framework for performing precise and comprehensive instance segmentation, enabling automatic preparation of concept sets from a given image. This paper for the first time explores using SAM to augment concept-based XAI. We offer an effective and flexible concept-based explanation method, namely Explain Any Concept (EAC), which explains DNN decisions with any concept. While SAM is highly effective and offers an "out-of-the-box" instance segmentation, it is costly when being integrated into defacto XAI pipelines. We thus propose a lightweight per-input equivalent (PIE) scheme, enabling efficient explanation with a surrogate model. Our evaluation over two popular datasets (ImageNet and COCO) illustrate the highly encouraging performance of EAC over commonly-used XAI methods.
Nowadays, recognition-synthesis-based methods have been quite popular with voice conversion (VC). By introducing linguistics features with good disentangling characters extracted from an automatic speech recognition (ASR) model, the VC performance achieved considerable breakthroughs. Recently, self-supervised learning (SSL) methods trained with a large-scale unannotated speech corpus have been applied to downstream tasks focusing on the content information, which is suitable for VC tasks. However, a huge amount of speaker information in SSL representations degrades timbre similarity and the quality of converted speech significantly. To address this problem, we proposed a high-similarity any-to-one voice conversion method with the input of SSL representations. We incorporated adversarial training mechanisms in the synthesis module using external unannotated corpora. Two auxiliary discriminators were trained to distinguish whether a sequence of mel-spectrograms has been converted by the acoustic model and whether a sequence of content embeddings contains speaker information from external corpora. Experimental results show that our proposed method achieves comparable similarity and higher naturalness than the supervised method, which needs a huge amount of annotated corpora for training and is applicable to improve similarity for VC methods with other SSL representations as input.