Transformers have recently dominated the ASR field. Although able to yield good performance, they involve an autoregressive (AR) decoder to generate tokens one by one, which is computationally inefficient. To speed up inference, non-autoregressive (NAR) methods, e.g. single-step NAR, were designed, to enable parallel generation. However, due to an independence assumption within the output tokens, performance of single-step NAR is inferior to that of AR models, especially with a large-scale corpus. There are two challenges to improving single-step NAR: Firstly to accurately predict the number of output tokens and extract hidden variables; secondly, to enhance modeling of interdependence between output tokens. To tackle both challenges, we propose a fast and accurate parallel transformer, termed Paraformer. This utilizes a continuous integrate-and-fire based predictor to predict the number of tokens and generate hidden variables. A glancing language model (GLM) sampler then generates semantic embeddings to enhance the NAR decoder's ability to model context interdependence. Finally, we design a strategy to generate negative samples for minimum word error rate training to further improve performance. Experiments using the public AISHELL-1, AISHELL-2 benchmark, and an industrial-level 20,000 hour task demonstrate that the proposed Paraformer can attain comparable performance to the state-of-the-art AR transformer, with more than 10x speedup.
Though the state-of-the architectures for semantic segmentation, such as HRNet, demonstrate impressive accuracy, the complexity arising from their salient design choices hinders a range of model acceleration tools, and further they make use of operations that are inefficient on current hardware. This paper demonstrates that a simple encoder-decoder architecture with a ResNet-like backbone and a small multi-scale head, performs on-par or better than complex semantic segmentation architectures such as HRNet, FANet and DDRNets. Naively applying deep backbones designed for Image Classification to the task of Semantic Segmentation leads to sub-par results, owing to a much smaller effective receptive field of these backbones. Implicit among the various design choices put forth in works like HRNet, DDRNet, and FANet are networks with a large effective receptive field. It is natural to ask if a simple encoder-decoder architecture would compare favorably if comprised of backbones that have a larger effective receptive field, though without the use of inefficient operations like dilated convolutions. We show that with minor and inexpensive modifications to ResNets, enlarging the receptive field, very simple and competitive baselines can be created for Semantic Segmentation. We present a family of such simple architectures for desktop as well as mobile targets, which match or exceed the performance of complex models on the Cityscapes dataset. We hope that our work provides simple yet effective baselines for practitioners to develop efficient semantic segmentation models.
Image captioning is the process of automatically generating a description of an image in natural language. Image captioning is one of the significant challenges in image understanding since it requires not only recognizing salient objects in the image but also their attributes and the way they interact. The system must then generate a syntactically and semantically correct caption that describes the image content in natural language. With the significant progress in deep learning models and their ability to effectively encode large sets of images and generate correct sentences, several neural-based captioning approaches have been proposed recently, each trying to achieve better accuracy and caption quality. This paper introduces an encoder-decoder-based image captioning system in which the encoder extracts spatial features from the image using ResNet-101. This stage is followed by a refining model, which uses an attention-on-attention mechanism to extract the visual features of the target image objects, then determine their interactions. The decoder consists of an attention-based recurrent module and a reflective attention module, which collaboratively apply attention to the visual and textual features to enhance the decoder's ability to model long-term sequential dependencies. Extensive experiments performed on Flickr30K, show the effectiveness of the proposed approach and the high quality of the generated captions.
NatiQ is end-to-end text-to-speech system for Arabic. Our speech synthesizer uses an encoder-decoder architecture with attention. We used both tacotron-based models (tacotron-1 and tacotron-2) and the faster transformer model for generating mel-spectrograms from characters. We concatenated Tacotron1 with the WaveRNN vocoder, Tacotron2 with the WaveGlow vocoder and ESPnet transformer with the parallel wavegan vocoder to synthesize waveforms from the spectrograms. We used in-house speech data for two voices: 1) neutral male "Hamza"- narrating general content and news, and 2) expressive female "Amina"- narrating children story books to train our models. Our best systems achieve an average Mean Opinion Score (MOS) of 4.21 and 4.40 for Amina and Hamza respectively. The objective evaluation of the systems using word and character error rate (WER and CER) as well as the response time measured by real-time factor favored the end-to-end architecture ESPnet. NatiQ demo is available on-line at https://tts.qcri.org
AMR-to-text is one of the key techniques in the NLP community that aims at generating sentences from the Abstract Meaning Representation (AMR) graphs. Since AMR was proposed in 2013, the study on AMR-to-Text has become increasingly prevalent as an essential branch of structured data to text because of the unique advantages of AMR as a high-level semantic description of natural language. In this paper, we provide a brief survey of AMR-to-Text. Firstly, we introduce the current scenario of this technique and point out its difficulties. Secondly, based on the methods used in previous studies, we roughly divided them into five categories according to their respective mechanisms, i.e., Rules-based, Seq-to-Seq-based, Graph-to-Seq-based, Transformer-based, and Pre-trained Language Model (PLM)-based. In particular, we detail the neural network-based method and present the latest progress of AMR-to-Text, which refers to AMR reconstruction, Decoder optimization, etc. Furthermore, we present the benchmarks and evaluation methods of AMR-to-Text. Eventually, we provide a summary of current techniques and the outlook for future research.
High-resolution synthesis/projection of images over a large field-of-view (FOV) is hindered by the restricted space-bandwidth-product (SBP) of wavefront modulators. We report a deep learning-enabled diffractive display design that is based on a jointly-trained pair of an electronic encoder and a diffractive optical decoder to synthesize/project super-resolved images using low-resolution wavefront modulators. The digital encoder, composed of a trained convolutional neural network (CNN), rapidly pre-processes the high-resolution images of interest so that their spatial information is encoded into low-resolution (LR) modulation patterns, projected via a low SBP wavefront modulator. The diffractive decoder processes this LR encoded information using thin transmissive layers that are structured using deep learning to all-optically synthesize and project super-resolved images at its output FOV. Our results indicate that this diffractive image display can achieve a super-resolution factor of ~4, demonstrating a ~16-fold increase in SBP. We also experimentally validate the success of this diffractive super-resolution display using 3D-printed diffractive decoders that operate at the THz spectrum. This diffractive image decoder can be scaled to operate at visible wavelengths and inspire the design of large FOV and high-resolution displays that are compact, low-power, and computationally efficient.
Unified vision-language frameworks have greatly advanced in recent years, most of which adopt an encoder-decoder architecture to unify image-text tasks as sequence-to-sequence generation. However, existing video-language (VidL) models still require task-specific designs in model architecture and training objectives for each task. In this work, we explore a unified VidL framework LAVENDER, where Masked Language Modeling (MLM) is used as the common interface for all pre-training and downstream tasks. Such unification leads to a simplified model architecture, where only a lightweight MLM head, instead of a decoder with much more parameters, is needed on top of the multimodal encoder. Surprisingly, experimental results show that this unified framework achieves competitive performance on 14 VidL benchmarks, covering video question answering, text-to-video retrieval and video captioning. Extensive analyses further demonstrate the advantage of LAVENDER over existing VidL methods in: (i) supporting all downstream tasks with just a single set of parameter values when multi-task finetuned; (ii) few-shot generalization on various downstream tasks; and (iii) enabling zero-shot evaluation on video question answering tasks. Code is available at https://github.com/microsoft/LAVENDER.
Federated learning is an emerging paradigm allowing large-scale decentralized learning without sharing data across different data owners, which helps address the concern of data privacy in medical image analysis. However, the requirement for label consistency across clients by the existing methods largely narrows its application scope. In practice, each clinical site may only annotate certain organs of interest with partial or no overlap with other sites. Incorporating such partially labeled data into a unified federation is an unexplored problem with clinical significance and urgency. This work tackles the challenge by using a novel federated multi-encoding U-Net (Fed-MENU) method for multi-organ segmentation. In our method, a multi-encoding U-Net (MENU-Net) is proposed to extract organ-specific features through different encoding sub-networks. Each sub-network can be seen as an expert of a specific organ and trained for that client. Moreover, to encourage the organ-specific features extracted by different sub-networks to be informative and distinctive, we regularize the training of the MENU-Net by designing an auxiliary generic decoder (AGD). Extensive experiments on six public abdominal CT datasets show that our Fed-MENU method can effectively obtain a federated learning model using the partially labeled datasets with superior performance to other models trained by either localized or centralized learning methods. Source code is publicly available at https://github.com/DIAL-RPI/Fed-MENU.
Neural approximations of scalar and vector fields, such as signed distance functions and radiance fields, have emerged as accurate, high-quality representations. State-of-the-art results are obtained by conditioning a neural approximation with a lookup from trainable feature grids that take on part of the learning task and allow for smaller, more efficient neural networks. Unfortunately, these feature grids usually come at the cost of significantly increased memory consumption compared to stand-alone neural network models. We present a dictionary method for compressing such feature grids, reducing their memory consumption by up to 100x and permitting a multiresolution representation which can be useful for out-of-core streaming. We formulate the dictionary optimization as a vector-quantized auto-decoder problem which lets us learn end-to-end discrete neural representations in a space where no direct supervision is available and with dynamic topology and structure. Our source code will be available at https://github.com/nv-tlabs/vqad.
We present a novel meta-learning approach for 6D pose estimation on unknown objects. In contrast to ``instance-level" and ``category-level" pose estimation methods, our algorithm learns object representation in a category-agnostic way, which endows it with strong generalization capabilities across object categories. Specifically, we employ a neural process-based meta-learning approach to train an encoder to capture texture and geometry of an object in a latent representation, based on very few RGB-D images and ground-truth keypoints. The latent representation is then used by a simultaneously meta-trained decoder to predict the 6D pose of the object in new images. Furthermore, we propose a novel geometry-aware decoder for the keypoint prediction using a Graph Neural Network (GNN), which explicitly takes geometric constraints specific to each object into consideration. To evaluate our algorithm, extensive experiments are conducted on the \linemod dataset, and on our new fully-annotated synthetic datasets generated from Multiple Categories in Multiple Scenes (MCMS). Experimental results demonstrate that our model performs well on unseen objects with very different shapes and appearances. Remarkably, our model also shows robust performance on occluded scenes although trained fully on data without occlusion. To our knowledge, this is the first work exploring \textbf{cross-category level} 6D pose estimation.
This paper considers the problem of undersampled MRI reconstruction. We propose a novel Transformer-based framework for directly processing signal in k-space, going beyond the limitation of regular grids as ConvNets do. We adopt an implicit representation of k-space spectrogram, treating spatial coordinates as inputs, and dynamically query the sparsely sampled points to reconstruct the spectrogram, i.e. learning the inductive bias in k-space. To strike a balance between computational cost and reconstruction quality, we build the decoder with hierarchical structure to generate low-resolution and high-resolution outputs respectively. To validate the effectiveness of our proposed method, we have conducted extensive experiments on two public datasets, and demonstrate superior or comparable performance to state-of-the-art approaches.
Vision transformers (ViTs) are changing the landscape of object detection approaches. A natural usage of ViTs in detection is to replace the CNN-based backbone with a transformer-based backbone, which is straightforward and effective, with the price of bringing considerable computation burden for inference. More subtle usage is the DETR family, which eliminates the need for many hand-designed components in object detection but introduces a decoder demanding an extra-long time to converge. As a result, transformer-based object detection can not prevail in large-scale applications. To overcome these issues, we propose a novel decoder-free fully transformer-based (DFFT) object detector, achieving high efficiency in both training and inference stages, for the first time. We simplify objection detection into an encoder-only single-level anchor-based dense prediction problem by centering around two entry points: 1) Eliminate the training-inefficient decoder and leverage two strong encoders to preserve the accuracy of single-level feature map prediction; 2) Explore low-level semantic features for the detection task with limited computational resources. In particular, we design a novel lightweight detection-oriented transformer backbone that efficiently captures low-level features with rich semantics based on a well-conceived ablation study. Extensive experiments on the MS COCO benchmark demonstrate that DFFT_SMALL outperforms DETR by 2.5% AP with 28% computation cost reduction and more than $10$x fewer training epochs. Compared with the cutting-edge anchor-based detector RetinaNet, DFFT_SMALL obtains over 5.5% AP gain while cutting down 70% computation cost.
This work studies the joint rain and haze removal problem. In real-life scenarios, rain and haze, two often co-occurring common weather phenomena, can greatly degrade the clarity and quality of the scene images, leading to a performance drop in the visual applications, such as autonomous driving. However, jointly removing the rain and haze in scene images is ill-posed and challenging, where the existence of haze and rain and the change of atmosphere light, can both degrade the scene information. Current methods focus on the contamination removal part, thus ignoring the restoration of the scene information affected by the change of atmospheric light. We propose a novel deep neural network, named Asymmetric Dual-decoder U-Net (ADU-Net), to address the aforementioned challenge. The ADU-Net produces both the contamination residual and the scene residual to efficiently remove the rain and haze while preserving the fidelity of the scene information. Extensive experiments show our work outperforms the existing state-of-the-art methods by a considerable margin in both synthetic data and real-world data benchmarks, including RainCityscapes, BID Rain, and SPA-Data. For instance, we improve the state-of-the-art PSNR value by 2.26/4.57 on the RainCityscapes/SPA-Data, respectively. Codes will be made available freely to the research community.
A compositional understanding of the world in terms of objects and their geometry in 3D space is considered a cornerstone of human cognition. Facilitating the learning of such a representation in neural networks holds promise for substantially improving labeled data efficiency. As a key step in this direction, we make progress on the problem of learning 3D-consistent decompositions of complex scenes into individual objects in an unsupervised fashion. We introduce Object Scene Representation Transformer (OSRT), a 3D-centric model in which individual object representations naturally emerge through novel view synthesis. OSRT scales to significantly more complex scenes with larger diversity of objects and backgrounds than existing methods. At the same time, it is multiple orders of magnitude faster at compositional rendering thanks to its light field parametrization and the novel Slot Mixer decoder. We believe this work will not only accelerate future architecture exploration and scaling efforts, but it will also serve as a useful tool for both object-centric as well as neural scene representation learning communities.
This paper describes the submission of our end-to-end YiTrans speech translation system for the IWSLT 2022 offline task, which translates from English audio to German, Chinese, and Japanese. The YiTrans system is built on large-scale pre-trained encoder-decoder models. More specifically, we first design a multi-stage pre-training strategy to build a multi-modality model with a large amount of labeled and unlabeled data. We then fine-tune the corresponding components of the model for the downstream speech translation tasks. Moreover, we make various efforts to improve performance, such as data filtering, data augmentation, speech segmentation, model ensemble, and so on. Experimental results show that our YiTrans system obtains a significant improvement than the strong baseline on three translation directions, and it achieves +5.2 BLEU improvements over last year's optimal end-to-end system on tst2021 English-German. Our final submissions rank first on English-German and English-Chinese end-to-end systems in terms of the automatic evaluation metric. We make our code and models publicly available.
An ideal music synthesizer should be both interactive and expressive, generating high-fidelity audio in realtime for arbitrary combinations of instruments and notes. Recent neural synthesizers have exhibited a tradeoff between domain-specific models that offer detailed control of only specific instruments, or raw waveform models that can train on any music but with minimal control and slow generation. In this work, we focus on a middle ground of neural synthesizers that can generate audio from MIDI sequences with arbitrary combinations of instruments in realtime. This enables training on a wide range of transcription datasets with a single model, which in turn offers note-level control of composition and instrumentation across a wide range of instruments. We use a simple two-stage process: MIDI to spectrograms with an encoder-decoder Transformer, then spectrograms to audio with a generative adversarial network (GAN) spectrogram inverter. We compare training the decoder as an autoregressive model and as a Denoising Diffusion Probabilistic Model (DDPM) and find that the DDPM approach is superior both qualitatively and as measured by audio reconstruction and Fr\'echet distance metrics. Given the interactivity and generality of this approach, we find this to be a promising first step towards interactive and expressive neural synthesis for arbitrary combinations of instruments and notes.
Common Neural Architecture Search methods generate large amounts of candidate architectures that need training in order to assess their performance and find an optimal architecture. To minimize the search time we use different performance estimation strategies. The effectiveness of such strategies varies in terms of accuracy and fit and query time. This study proposes a new method, EmProx Score (Embedding Proximity Score). Similar to Neural Architecture Optimization (NAO), this method maps candidate architectures to a continuous embedding space using an encoder-decoder framework. The performance of candidates is then estimated using weighted kNN based on the embedding vectors of architectures of which the performance is known. Performance estimations of this method are comparable to the MLP performance predictor used in NAO in terms of accuracy, while being nearly nine times faster to train compared to NAO. Benchmarking against other performance estimation strategies currently used shows similar to better accuracy, while being five up to eighty times faster.
Clinically, the accurate annotation of lesions/tissues can significantly facilitate the disease diagnosis. For example, the segmentation of optic disc/cup (OD/OC) on fundus image would facilitate the glaucoma diagnosis, the segmentation of skin lesions on dermoscopic images is helpful to the melanoma diagnosis, etc. With the advancement of deep learning techniques, a wide range of methods proved the lesions/tissues segmentation can also facilitate the automated disease diagnosis models. However, existing methods are limited in the sense that they can only capture static regional correlations in the images. Inspired by the global and dynamic nature of Vision Transformer, in this paper, we propose Segmentation-Assisted diagnosis Transformer (SeATrans) to transfer the segmentation knowledge to the disease diagnosis network. Specifically, we first propose an asymmetric multi-scale interaction strategy to correlate each single low-level diagnosis feature with multi-scale segmentation features. Then, an effective strategy called SeA-block is adopted to vitalize diagnosis feature via correlated segmentation features. To model the segmentation-diagnosis interaction, SeA-block first embeds the diagnosis feature based on the segmentation information via the encoder, and then transfers the embedding back to the diagnosis feature space by a decoder. Experimental results demonstrate that SeATrans surpasses a wide range of state-of-the-art (SOTA) segmentation-assisted diagnosis methods on several disease diagnosis tasks.
The detection and removal of precancerous polyps through colonoscopy is the primary technique for the prevention of colorectal cancer worldwide. However, the miss rate of colorectal polyp varies significantly among the endoscopists. It is well known that a computer-aided diagnosis (CAD) system can assist endoscopists in detecting colon polyps and minimize the variation among endoscopists. In this study, we introduce a novel deep learning architecture, named MKDCNet, for automatic polyp segmentation robust to significant changes in polyp data distribution. MKDCNet is simply an encoder-decoder neural network that uses the pre-trained ResNet50 as the encoder and novel multiple kernel dilated convolution (MKDC) block that expands the field of view to learn more robust and heterogeneous representation. Extensive experiments on four publicly available polyp datasets and cell nuclei dataset show that the proposed MKDCNet outperforms the state-of-the-art methods when trained and tested on the same dataset as well when tested on unseen polyp datasets from different distributions. With rich results, we demonstrated the robustness of the proposed architecture. From an efficiency perspective, our algorithm can process at (approx 45) frames per second on RTX 3090 GPU. MKDCNet can be a strong benchmark for building real-time systems for clinical colonoscopies. The code of the proposed MKDCNet is available at https://github.com/nikhilroxtomar/MKDCNet.
Time-series forecasting plays an important role in many real-world scenarios, such as equipment life cycle forecasting, weather forecasting, and traffic flow forecasting. It can be observed from recent research that a variety of transformer-based models have shown remarkable results in time-series forecasting. However, there are still some issues that limit the ability of transformer-based models on time-series forecasting tasks: (i) learning directly on raw data is susceptible to noise due to its complex and unstable feature representation; (ii) the self-attention mechanisms pay insufficient attention to changing features and temporal dependencies. In order to solve these two problems, we propose a transformer-based differentially reconstructed attention model DRAformer. Specifically, DRAformer has the following innovations: (i) learning against differenced sequences, which preserves clear and stable sequence features by differencing and highlights the changing properties of sequences; (ii) the reconstructed attention: integrated distance attention exhibits sequential distance through a learnable Gaussian kernel, distributed difference attention calculates distribution difference by mapping the difference sequence to the adaptive feature space, and the combination of the two effectively focuses on the sequences with prominent associations; (iii) the reconstructed decoder input, which extracts sequence features by integrating variation information and temporal correlations, thereby obtaining a more comprehensive sequence representation. Extensive experiments on four large-scale datasets demonstrate that DRAformer outperforms state-of-the-art baselines.