Deep learning has transformed computational imaging, but traditional pixel-based representations limit their ability to capture continuous, multiscale details of objects. Here we introduce a novel Local Conditional Neural Fields (LCNF) framework, leveraging a continuous implicit neural representation to address this limitation. LCNF enables flexible object representation and facilitates the reconstruction of multiscale information. We demonstrate the capabilities of LCNF in solving the highly ill-posed inverse problem in Fourier ptychographic microscopy (FPM) with multiplexed measurements, achieving robust, scalable, and generalizable large-scale phase retrieval. Unlike traditional neural fields frameworks, LCNF incorporates a local conditional representation that promotes model generalization, learning multiscale information, and efficient processing of large-scale imaging data. By combining an encoder and a decoder conditioned on a learned latent vector, LCNF achieves versatile continuous-domain super-resolution image reconstruction. We demonstrate accurate reconstruction of wide field-of-view, high-resolution phase images using only a few multiplexed measurements. LCNF robustly captures the continuous object priors and eliminates various phase artifacts, even when it is trained on imperfect datasets. The framework exhibits strong generalization, reconstructing diverse objects even with limited training data. Furthermore, LCNF can be trained on a physics simulator using natural images and successfully applied to experimental measurements on biological samples. Our results highlight the potential of LCNF for solving large-scale inverse problems in computational imaging, with broad applicability in various deep-learning-based techniques.
Most real-world 3D measurements from depth sensors are incomplete, and to address this issue the point cloud completion task aims to predict the complete shapes of objects from partial observations. Previous works often adapt an encoder-decoder architecture, where the encoder is trained to extract embeddings that are used as inputs to generate predictions from the decoder. However, the learned embeddings have sparse distribution in the feature space, which leads to worse generalization results during testing. To address these problems, this paper proposes a hyperspherical module, which transforms and normalizes embeddings from the encoder to be on a unit hypersphere. With the proposed module, the magnitude and direction of the output hyperspherical embedding are decoupled and only the directional information is optimized. We theoretically analyze the hyperspherical embedding and show that it enables more stable training with a wider range of learning rates and more compact embedding distributions. Experiment results show consistent improvement of point cloud completion in both single-task and multi-task learning, which demonstrates the effectiveness of the proposed method.
Automatic tree density estimation and counting using single aerial and satellite images is a challenging task in photogrammetry and remote sensing, yet has an important role in forest management. In this paper, we propose the first semisupervised transformer-based framework for tree counting which reduces the expensive tree annotations for remote sensing images. Our method, termed as TreeFormer, first develops a pyramid tree representation module based on transformer blocks to extract multi-scale features during the encoding stage. Contextual attention-based feature fusion and tree density regressor modules are further designed to utilize the robust features from the encoder to estimate tree density maps in the decoder. Moreover, we propose a pyramid learning strategy that includes local tree density consistency and local tree count ranking losses to utilize unlabeled images into the training process. Finally, the tree counter token is introduced to regulate the network by computing the global tree counts for both labeled and unlabeled images. Our model was evaluated on two benchmark tree counting datasets, Jiangsu, and Yosemite, as well as a new dataset, KCL-London, created by ourselves. Our TreeFormer outperforms the state of the art semi-supervised methods under the same setting and exceeds the fully-supervised methods using the same number of labeled images. The codes and datasets are available at https://github.com/HAAClassic/TreeFormer.
Semantic image synthesis (SIS) refers to the problem of generating realistic imagery given a semantic segmentation mask that defines the spatial layout of object classes. Most of the approaches in the literature, other than the quality of the generated images, put effort in finding solutions to increase the generation diversity in terms of style i.e. texture. However, they all neglect a different feature, which is the possibility of manipulating the layout provided by the mask. Currently, the only way to do so is manually by means of graphical users interfaces. In this paper, we describe a network architecture to address the problem of automatically manipulating or generating the shape of object classes in semantic segmentation masks, with specific focus on human faces. Our proposed model allows embedding the mask class-wise into a latent space where each class embedding can be independently edited. Then, a bi-directional LSTM block and a convolutional decoder output a new, locally manipulated mask. We report quantitative and qualitative results on the CelebMask-HQ dataset, which show our model can both faithfully reconstruct and modify a segmentation mask at the class level. Also, we show our model can be put before a SIS generator, opening the way to a fully automatic generation control of both shape and texture. Code available at https://github.com/TFonta/Semantic-VAE.
Taking long-term spectral and temporal dependencies into account is essential for automatic piano transcription. This is especially helpful when determining the precise onset and offset for each note in the polyphonic piano content. In this case, we may rely on the capability of self-attention mechanism in Transformers to capture these long-term dependencies in the frequency and time axes. In this work, we propose hFT-Transformer, which is an automatic music transcription method that uses a two-level hierarchical frequency-time Transformer architecture. The first hierarchy includes a convolutional block in the time axis, a Transformer encoder in the frequency axis, and a Transformer decoder that converts the dimension in the frequency axis. The output is then fed into the second hierarchy which consists of another Transformer encoder in the time axis. We evaluated our method with the widely used MAPS and MAESTRO v3.0.0 datasets, and it demonstrated state-of-the-art performance on all the F1-scores of the metrics among Frame, Note, Note with Offset, and Note with Offset and Velocity estimations.
Survival prediction is crucial for cancer patients as it provides early prognostic information for treatment planning. Recently, deep survival models based on deep learning and medical images have shown promising performance for survival prediction. However, existing deep survival models are not well developed in utilizing multi-modality images (e.g., PET-CT) and in extracting region-specific information (e.g., the prognostic information in Primary Tumor (PT) and Metastatic Lymph Node (MLN) regions). In view of this, we propose a merging-diverging learning framework for survival prediction from multi-modality images. This framework has a merging encoder to fuse multi-modality information and a diverging decoder to extract region-specific information. In the merging encoder, we propose a Hybrid Parallel Cross-Attention (HPCA) block to effectively fuse multi-modality features via parallel convolutional layers and cross-attention transformers. In the diverging decoder, we propose a Region-specific Attention Gate (RAG) block to screen out the features related to lesion regions. Our framework is demonstrated on survival prediction from PET-CT images in Head and Neck (H&N) cancer, by designing an X-shape merging-diverging hybrid transformer network (named XSurv). Our XSurv combines the complementary information in PET and CT images and extracts the region-specific prognostic information in PT and MLN regions. Extensive experiments on the public dataset of HEad and neCK TumOR segmentation and outcome prediction challenge (HECKTOR 2022) demonstrate that our XSurv outperforms state-of-the-art survival prediction methods.
Signed graphs are complex systems that represent trust relationships or preferences in various domains. Learning node representations in such graphs is crucial for many mining tasks. Although real-world signed relationships can be influenced by multiple latent factors, most existing methods often oversimplify the modeling of signed relationships by relying on social theories and treating them as simplistic factors. This limits their expressiveness and their ability to capture the diverse factors that shape these relationships. In this paper, we propose DINES, a novel method for learning disentangled node representations in signed directed graphs without social assumptions. We adopt a disentangled framework that separates each embedding into distinct factors, allowing for capturing multiple latent factors. We also explore lightweight graph convolutions that focus solely on sign and direction, without depending on social theories. Additionally, we propose a decoder that effectively classifies an edge's sign by considering correlations between the factors. To further enhance disentanglement, we jointly train a self-supervised factor discriminator with our encoder and decoder. Throughout extensive experiments on real-world signed directed graphs, we show that DINES effectively learns disentangled node representations, and significantly outperforms its competitors in the sign prediction task.
Deep learning has become a valuable tool for the automation of certain medical image segmentation tasks, significantly relieving the workload of medical specialists. Some of these tasks require segmentation to be performed on a subset of the input dimensions, the most common case being 3D-to-2D. However, the performance of existing methods is strongly conditioned by the amount of labeled data available, as there is currently no data efficient method, e.g. transfer learning, that has been validated on these tasks. In this work, we propose a novel convolutional neural network (CNN) and self-supervised learning (SSL) method for label-efficient 3D-to-2D segmentation. The CNN is composed of a 3D encoder and a 2D decoder connected by novel 3D-to-2D blocks. The SSL method consists of reconstructing image pairs of modalities with different dimensionality. The approach has been validated in two tasks with clinical relevance: the en-face segmentation of geographic atrophy and reticular pseudodrusen in optical coherence tomography. Results on different datasets demonstrate that the proposed CNN significantly improves the state of the art in scenarios with limited labeled data by up to 8% in Dice score. Moreover, the proposed SSL method allows further improvement of this performance by up to 23%, and we show that the SSL is beneficial regardless of the network architecture.
U-Net style networks are commonly utilized in unsupervised image registration to predict dense displacement fields, which for high-resolution volumetric image data is a resource-intensive and time-consuming task. To tackle this challenge, we first propose Fourier-Net, which replaces the costly U-Net style expansive path with a parameter-free model-driven decoder. Instead of directly predicting a full-resolution displacement field, our Fourier-Net learns a low-dimensional representation of the displacement field in the band-limited Fourier domain which our model-driven decoder converts to a full-resolution displacement field in the spatial domain. Expanding upon Fourier-Net, we then introduce Fourier-Net+, which additionally takes the band-limited spatial representation of the images as input and further reduces the number of convolutional layers in the U-Net style network's contracting path. Finally, to enhance the registration performance, we propose a cascaded version of Fourier-Net+. We evaluate our proposed methods on three datasets, on which our proposed Fourier-Net and its variants achieve comparable results with current state-of-the art methods, while exhibiting faster inference speeds, lower memory footprint, and fewer multiply-add operations. With such small computational cost, our Fourier-Net+ enables the efficient training of large-scale 3D registration on low-VRAM GPUs. Our code is publicly available at \url{https://github.com/xi-jia/Fourier-Net}.
Although unsupervised domain adaptation (UDA) is a promising direction to alleviate domain shift, they fall short of their supervised counterparts. In this work, we investigate relatively less explored semi-supervised domain adaptation (SSDA) for medical image segmentation, where access to a few labeled target samples can improve the adaptation performance substantially. Specifically, we propose a two-stage training process. First, an encoder is pre-trained in a self-learning paradigm using a novel domain-content disentangled contrastive learning (CL) along with a pixel-level feature consistency constraint. The proposed CL enforces the encoder to learn discriminative content-specific but domain-invariant semantics on a global scale from the source and target images, whereas consistency regularization enforces the mining of local pixel-level information by maintaining spatial sensitivity. This pre-trained encoder, along with a decoder, is further fine-tuned for the downstream task, (i.e. pixel-level segmentation) using a semi-supervised setting. Furthermore, we experimentally validate that our proposed method can easily be extended for UDA settings, adding to the superiority of the proposed strategy. Upon evaluation on two domain adaptive image segmentation tasks, our proposed method outperforms the SoTA methods, both in SSDA and UDA settings. Code is available at https://github.com/hritam-98/GFDA-disentangled
Recently, the release of INSTRUCTEVAL has provided valuable insights into the performance of large language models (LLMs) that utilize encoder-decoder or decoder-only architecture. Interestingly, despite being introduced four years ago, T5-based LLMs, such as FLAN-T5, continue to outperform the latest decoder-based LLMs, such as LLAMA and VICUNA, on tasks that require general problem-solving skills. This performance discrepancy can be attributed to three key factors: (1) Pre-training data, (2) Backbone architecture, and (3) Instruction dataset. In this technical report, our main focus is on investigating the impact of the third factor by leveraging VICUNA, a large language model based on LLAMA, which has undergone fine-tuning on ChatGPT conversations. To achieve this objective, we fine-tuned VICUNA using a customized instruction dataset collection called FLANMINI. This collection includes a subset of the large-scale instruction dataset known as FLAN, as well as various code-related datasets and conversational datasets derived from ChatGPT/GPT-4. This dataset comprises a large number of tasks that demand problem-solving skills. Our experimental findings strongly indicate that the enhanced problem-solving abilities of our model, FLACUNA, are obtained through fine-tuning VICUNA on the FLAN dataset, leading to significant improvements across numerous benchmark datasets in INSTRUCTEVAL. FLACUNA is publicly available at https://huggingface.co/declare-lab/flacuna-13b-v1.0.
Protoform reconstruction is the task of inferring what morphemes or words appeared like in the ancestral languages of a set of daughter languages. Meloni et al. (2021) achieved the state-of-the-art on Latin protoform reconstruction with an RNN-based encoder-decoder with attention model. We update their model with the state-of-the-art seq2seq model: the Transformer. Our model outperforms their model on a suite of different metrics on two different datasets: their Romance data of 8,000 cognates spanning 5 languages and a Chinese dataset (Hou 2004) of 800+ cognates spanning 39 varieties. We also probe our model for potential phylogenetic signal contained in the model. Our code is publicly available at https://github.com/cmu-llab/acl-2023.
In this report, we describe the technical details of our approach for the Ego4D Long-Term Action Anticipation Challenge 2023. The aim of this task is to predict a sequence of future actions that will take place at an arbitrary time or later, given an input video. To accomplish this task, we introduce three improvements to the baseline model, which consists of an encoder that generates clip-level features from the video, an aggregator that integrates multiple clip-level features, and a decoder that outputs Z future actions. 1) Model ensemble of SlowFast and SlowFast-CLIP; 2) Label smoothing to relax order constraints for future actions; 3) Constraining the prediction of the action class (verb, noun) based on word co-occurrence. Our method outperformed the baseline performance and recorded as second place solution on the public leaderboard.
Limited by imaging systems, the reconstruction of Magnetic Resonance Imaging (MRI) images from partial measurement is essential to medical imaging research. Benefiting from the diverse and complementary information of multi-contrast MR images in different imaging modalities, multi-contrast Super-Resolution (SR) reconstruction is promising to yield SR images with higher quality. In the medical scenario, to fully visualize the lesion, radiologists are accustomed to zooming the MR images at arbitrary scales rather than using a fixed scale, as used by most MRI SR methods. In addition, existing multi-contrast MRI SR methods often require a fixed resolution for the reference image, which makes acquiring reference images difficult and imposes limitations on arbitrary scale SR tasks. To address these issues, we proposed an implicit neural representations based dual-arbitrary multi-contrast MRI super-resolution method, called Dual-ArbNet. First, we decouple the resolution of the target and reference images by a feature encoder, enabling the network to input target and reference images at arbitrary scales. Then, an implicit fusion decoder fuses the multi-contrast features and uses an Implicit Decoding Function~(IDF) to obtain the final MRI SR results. Furthermore, we introduce a curriculum learning strategy to train our network, which improves the generalization and performance of our Dual-ArbNet. Extensive experiments in two public MRI datasets demonstrate that our method outperforms state-of-the-art approaches under different scale factors and has great potential in clinical practice.
Recent transformer-based methods achieve notable gains in the Human-object Interaction Detection (HOID) task by leveraging the detection of DETR and the prior knowledge of Vision-Language Model (VLM). However, these methods suffer from extended training times and complex optimization due to the entanglement of object detection and HOI recognition during the decoding process. Especially, the query embeddings used to predict both labels and boxes suffer from ambiguous representations, and the gap between the prediction of HOI labels and verb labels is not considered. To address these challenges, we introduce SOV-STG-VLA with three key components: Subject-Object-Verb (SOV) decoding, Specific Target Guided (STG) denoising, and a Vision-Language Advisor (VLA). Our SOV decoders disentangle object detection and verb recognition with a novel interaction region representation. The STG denoising strategy learns label embeddings with ground-truth information to guide the training and inference. Our SOV-STG achieves a fast convergence speed and high accuracy and builds a foundation for the VLA to incorporate the prior knowledge of the VLM. We introduce a vision advisor decoder to fuse both the interaction region information and the VLM's vision knowledge and a Verb-HOI prediction bridge to promote interaction representation learning. Our VLA notably improves our SOV-STG and achieves SOTA performance with one-sixth of training epochs compared to recent SOTA. Code and models are available at https://github.com/cjw2021/SOV-STG-VLA
Sequence modeling approaches have shown promising results in robot imitation learning. Recently, diffusion models have been adopted for behavioral cloning in a sequence modeling fashion, benefiting from their exceptional capabilities in modeling complex data distributions. The standard diffusion-based policy iteratively generates action sequences from random noise conditioned on the input states. Nonetheless, the model for diffusion policy can be further improved in terms of visual representations. In this work, we propose Crossway Diffusion, a simple yet effective method to enhance diffusion-based visuomotor policy learning via a carefully designed state decoder and an auxiliary self-supervised learning (SSL) objective. The state decoder reconstructs raw image pixels and other state information from the intermediate representations of the reverse diffusion process. The whole model is jointly optimized by the SSL objective and the original diffusion loss. Our experiments demonstrate the effectiveness of Crossway Diffusion in various simulated and real-world robot tasks, confirming its consistent advantages over the standard diffusion-based policy and substantial improvements over the baselines.
Unlocking the potential of deep learning in Peak-Hour Series Forecasting (PHSF) remains a critical yet underexplored task in various domains. While state-of-the-art deep learning models excel in regular Time Series Forecasting (TSF), they struggle to achieve comparable results in PHSF. This can be attributed to the challenges posed by the high degree of non-stationarity in peak-hour series, which makes direct forecasting more difficult than standard TSF. Additionally, manually extracting the maximum value from regular forecasting results leads to suboptimal performance due to models minimizing the mean deficit. To address these issues, this paper presents Seq2Peak, a novel framework designed specifically for PHSF tasks, bridging the performance gap observed in TSF models. Seq2Peak offers two key components: the CyclicNorm pipeline to mitigate the non-stationarity issue and a simple yet effective trainable-parameter-free peak-hour decoder with a hybrid loss function that utilizes both the original series and peak-hour series as supervised signals. Extensive experimentation on publicly available time series datasets demonstrates the effectiveness of the proposed framework, yielding a remarkable average relative improvement of 37.7% across four real-world datasets for both transformer- and non-transformer-based TSF models.
Domain-generalized urban-scene semantic segmentation (USSS) aims to learn generalized semantic predictions across diverse urban-scene styles. Unlike domain gap challenges, USSS is unique in that the semantic categories are often similar in different urban scenes, while the styles can vary significantly due to changes in urban landscapes, weather conditions, lighting, and other factors. Existing approaches typically rely on convolutional neural networks (CNNs) to learn the content of urban scenes. In this paper, we propose a Content-enhanced Mask TransFormer (CMFormer) for domain-generalized USSS. The main idea is to enhance the focus of the fundamental component, the mask attention mechanism, in Transformer segmentation models on content information. To achieve this, we introduce a novel content-enhanced mask attention mechanism. It learns mask queries from both the image feature and its down-sampled counterpart, as lower-resolution image features usually contain more robust content information and are less sensitive to style variations. These features are fused into a Transformer decoder and integrated into a multi-resolution content-enhanced mask attention learning scheme. Extensive experiments conducted on various domain-generalized urban-scene segmentation datasets demonstrate that the proposed CMFormer significantly outperforms existing CNN-based methods for domain-generalized semantic segmentation, achieving improvements of up to 14.00\% in terms of mIoU (mean intersection over union). The source code is publicly available at \url{https://github.com/BiQiWHU/CMFormer}.
Filter pruning simultaneously accelerates the computation and reduces the memory overhead of CNNs, which can be effectively applied to edge devices and cloud services. In this paper, we propose a novel Knowledge-driven Differential Filter Sampler~(KDFS) with Masked Filter Modeling~(MFM) framework for filter pruning, which globally prunes the redundant filters based on the prior knowledge of a pre-trained model in a differential and non-alternative optimization. Specifically, we design a differential sampler with learnable sampling parameters to build a binary mask vector for each layer, determining whether the corresponding filters are redundant. To learn the mask, we introduce masked filter modeling to construct PCA-like knowledge by aligning the intermediate features from the pre-trained teacher model and the outputs of the student decoder taking sampling features as the input. The mask and sampler are directly optimized by the Gumbel-Softmax Straight-Through Gradient Estimator in an end-to-end manner in combination with global pruning constraint, MFM reconstruction error, and dark knowledge. Extensive experiments demonstrate the proposed KDFS's effectiveness in compressing the base models on various datasets. For instance, the pruned ResNet-50 on ImageNet achieves $55.36\%$ computation reduction, and $42.86\%$ parameter reduction, while only dropping $0.35\%$ Top-1 accuracy, significantly outperforming the state-of-the-art methods. The code is available at \url{https://github.com/Osilly/KDFS}.
Motion prediction is crucial for autonomous driving systems to understand complex driving scenarios and make informed decisions. However, this task is challenging due to the diverse behaviors of traffic participants and complex environmental contexts. In this paper, we propose Motion TRansformer (MTR) frameworks to address these challenges. The initial MTR framework utilizes a transformer encoder-decoder structure with learnable intention queries, enabling efficient and accurate prediction of future trajectories. By customizing intention queries for distinct motion modalities, MTR improves multimodal motion prediction while reducing reliance on dense goal candidates. The framework comprises two essential processes: global intention localization, identifying the agent's intent to enhance overall efficiency, and local movement refinement, adaptively refining predicted trajectories for improved accuracy. Moreover, we introduce an advanced MTR++ framework, extending the capability of MTR to simultaneously predict multimodal motion for multiple agents. MTR++ incorporates symmetric context modeling and mutually-guided intention querying modules to facilitate future behavior interaction among multiple agents, resulting in scene-compliant future trajectories. Extensive experimental results demonstrate that the MTR framework achieves state-of-the-art performance on the highly-competitive motion prediction benchmarks, while the MTR++ framework surpasses its precursor, exhibiting enhanced performance and efficiency in predicting accurate multimodal future trajectories for multiple agents.