Spatio-Temporal Outdoor Lighting Aggregation on Image Sequences using Transformer Networks

In this work, we focus on outdoor lighting estimation by aggregating individual noisy estimates from images, exploiting the rich image information from wide-angle cameras and/or temporal image sequences. Photographs inherently encode information about the scene's lighting in the form of shading and shadows. Recovering the lighting is an inverse rendering problem and as that ill-posed. Recent work based on deep neural networks has shown promising results for single image lighting estimation, but suffers from robustness. We tackle this problem by combining lighting estimates from several image views sampled in the angular and temporal domain of an image sequence. For this task, we introduce a transformer architecture that is trained in an end-2-end fashion without any statistical post-processing as required by previous work. Thereby, we propose a positional encoding that takes into account the camera calibration and ego-motion estimation to globally register the individual estimates when computing attention between visual words. We show that our method leads to improved lighting estimation while requiring less hyper-parameters compared to the state-of-the-art.

Pretraining with Artificial Language: Studying Transferable Knowledge in Language Models

We investigate what kind of structural knowledge learned in neural network encoders is transferable to processing natural language. We design artificial languages with structural properties that mimic natural language, pretrain encoders on the data, and see how much performance the encoder exhibits on downstream tasks in natural language. Our experimental results show that pretraining with an artificial language with a nesting dependency structure provides some knowledge transferable to natural language. A follow-up probing analysis indicates that its success in the transfer is related to the amount of encoded contextual information and what is transferred is the knowledge of position-aware context dependence of language. Our results provide insights into how neural network encoders process human languages and the source of cross-lingual transferability of recent multilingual language models.

An Overview & Analysis of Sequence-to-Sequence Emotional Voice Conversion

Emotional voice conversion (EVC) focuses on converting a speech utterance from a source to a target emotion; it can thus be a key enabling technology for human-computer interaction applications and beyond. However, EVC remains an unsolved research problem with several challenges. In particular, as speech rate and rhythm are two key factors of emotional conversion, models have to generate output sequences of differing length. Sequence-to-sequence modelling is recently emerging as a competitive paradigm for models that can overcome those challenges. In an attempt to stimulate further research in this promising new direction, recent sequence-to-sequence EVC papers were systematically investigated and reviewed from six perspectives: their motivation, training strategies, model architectures, datasets, model inputs, and evaluation methods. This information is organised to provide the research community with an easily digestible overview of the current state-of-the-art. Finally, we discuss existing challenges of sequence-to-sequence EVC.

Parameter-Efficient Abstractive Question Answering over Tables or Text

A long-term ambition of information seeking QA systems is to reason over multi-modal contexts and generate natural answers to user queries. Today, memory intensive pre-trained language models are adapted to downstream tasks such as QA by fine-tuning the model on QA data in a specific modality like unstructured text or structured tables. To avoid training such memory-hungry models while utilizing a uniform architecture for each modality, parameter-efficient adapters add and train small task-specific bottle-neck layers between transformer layers. In this work, we study parameter-efficient abstractive QA in encoder-decoder models over structured tabular data and unstructured textual data using only 1.5% additional parameters for each modality. We also ablate over adapter layers in both encoder and decoder modules to study the efficiency-performance trade-off and demonstrate that reducing additional trainable parameters down to 0.7%-1.0% leads to comparable results. Our models out-perform current state-of-the-art models on tabular QA datasets such as Tablesum and FeTaQA, and achieve comparable performance on a textual QA dataset such as NarrativeQA using significantly less trainable parameters than fine-tuning.

Adversarial Defense via Image Denoising with Chaotic Encryption

In the literature on adversarial examples, white box and black box attacks have received the most attention. The adversary is assumed to have either full (white) or no (black) access to the defender's model. In this work, we focus on the equally practical gray box setting, assuming an attacker has partial information. We propose a novel defense that assumes everything but a private key will be made available to the attacker. Our framework uses an image denoising procedure coupled with encryption via a discretized Baker map. Extensive testing against adversarial images (e.g. FGSM, PGD) crafted using various gradients shows that our defense achieves significantly better results on CIFAR-10 and CIFAR-100 than the state-of-the-art gray box defenses in both natural and adversarial accuracy.

Few-shot Structured Radiology Report Generation Using Natural Language Prompts

Chest radiograph reporting is time-consuming, and numerous solutions to automate this process have been proposed. Due to the complexity of medical information, the variety of writing styles, and free text being prone to typos and inconsistencies, the efficacy of quantifying the clinical accuracy of free-text reports using natural language processing measures is challenging. On the other hand, structured reports ensure consistency and can more easily be used as a quality assurance tool. To accomplish this, we present a strategy for predicting clinical observations and their anatomical location that is easily extensible to other structured findings. First, we train a contrastive language-image model using related chest radiographs and free-text radiological reports. Then, we create textual prompts for each structured finding and optimize a classifier for predicting clinical findings and their associations within the medical image. The results indicate that even when only a few image-level annotations are used for training, the method can localize pathologies in chest radiographs and generate structured reports.

Transformer Inertial Poser: Attention-based Real-time Human Motion Reconstruction from Sparse IMUs

Real-time human motion reconstruction from a sparse set of wearable IMUs provides an non-intrusive and economic approach to motion capture. Without the ability to acquire absolute position information using IMUs, many prior works took data-driven approaches that utilize large human motion datasets to tackle the under-determined nature of the problem. Still, challenges such as temporal consistency, global translation estimation, and diverse coverage of motion or terrain types remain. Inspired by recent success of Transformer models in sequence modeling, we propose an attention-based deep learning method to reconstruct full-body motion from six IMU sensors in real-time. Together with a physics-based learning objective to predict "stationary body points", our method achieves new state-of-the-art results both quantitatively and qualitatively, while being simple to implement and smaller in size. We evaluate our method extensively on synthesized and real IMU data, and with real-time live demos.

DisARM: Displacement Aware Relation Module for 3D Detection

We introduce Displacement Aware Relation Module (DisARM), a novel neural network module for enhancing the performance of 3D object detection in point cloud scenes. The core idea of our method is that contextual information is critical to tell the difference when the instance geometry is incomplete or featureless. We find that relations between proposals provide a good representation to describe the context. However, adopting relations between all the object or patch proposals for detection is inefficient, and an imbalanced combination of local and global relations brings extra noise that could mislead the training. Rather than working with all relations, we found that training with relations only between the most representative ones, or anchors, can significantly boost the detection performance. A good anchor should be semantic-aware with no ambiguity and independent with other anchors as well. To find the anchors, we first perform a preliminary relation anchor module with an objectness-aware sampling approach and then devise a displacement-based module for weighing the relation importance for better utilization of contextual information. This lightweight relation module leads to significantly higher accuracy of object instance detection when being plugged into the state-of-the-art detectors. Evaluations on the public benchmarks of real-world scenes show that our method achieves state-of-the-art performance on both SUN RGB-D and ScanNet V2.

Integrating Contrastive Learning with Dynamic Models for Reinforcement Learning from Images

Recent methods for reinforcement learning from images use auxiliary tasks to learn image features that are used by the agent's policy or Q-function. In particular, methods based on contrastive learning that induce linearity of the latent dynamics or invariance to data augmentation have been shown to greatly improve the sample efficiency of the reinforcement learning algorithm and the generalizability of the learned embedding. We further argue, that explicitly improving Markovianity of the learned embedding is desirable and propose a self-supervised representation learning method which integrates contrastive learning with dynamic models to synergistically combine these three objectives: (1) We maximize the InfoNCE bound on the mutual information between the state- and action-embedding and the embedding of the next state to induce a linearly predictive embedding without explicitly learning a linear transition model, (2) we further improve Markovianity of the learned embedding by explicitly learning a non-linear transition model using regression, and (3) we maximize the mutual information between the two nonlinear predictions of the next embeddings based on the current action and two independent augmentations of the current state, which naturally induces transformation invariance not only for the state embedding, but also for the nonlinear transition model. Experimental evaluation on the Deepmind control suite shows that our proposed method achieves higher sample efficiency and better generalization than state-of-art methods based on contrastive learning or reconstruction.

Addressing Non-Intervention Challenges via Resilient Robotics utilizing a Digital Twin

Multi-robot systems face challenges in reducing human interventions as they are often deployed in dangerous environments. It is therefore necessary to include a methodology to assess robot failure rates to reduce the requirement for costly human intervention. A solution to this problem includes robots with the ability to work together to ensure mission resilience. To prevent this intervention, robots should be able to work together to ensure mission resilience. However, robotic platforms generally lack built-in interconnectivity with other platforms from different vendors. This work aims to tackle this issue by enabling the functionality through a bidirectional digital twin. The twin enables the human operator to transmit and receive information to and from the multi-robot fleet. This digital twin considers mission resilience, decision making and a run-time reliability ontology for failure detection to enable the resilience of a multi-robot fleet. This creates the cooperation, corroboration, and collaboration of diverse robots to leverage the capability of robots and support recovery of a failed robot.