Despite many attempts to leverage pre-trained text-to-image models (T2I) like Stable Diffusion (SD) for controllable image editing, producing good predictable results remains a challenge. Previous approaches have focused on either fine-tuning pre-trained T2I models on specific datasets to generate certain kinds of images (e.g., with a specific object or person), or on optimizing the weights, text prompts, and/or learning features for each input image in an attempt to coax the image generator to produce the desired result. However, these approaches all have shortcomings and fail to produce good results in a predictable and controllable manner. To address this problem, we present TiNO-Edit, an SD-based method that focuses on optimizing the noise patterns and diffusion timesteps during editing, something previously unexplored in the literature. With this simple change, we are able to generate results that both better align with the original images and reflect the desired result. Furthermore, we propose a set of new loss functions that operate in the latent domain of SD, greatly speeding up the optimization when compared to prior approaches, which operate in the pixel domain. Our method can be easily applied to variations of SD including Textual Inversion and DreamBooth that encode new concepts and incorporate them into the edited results. We present a host of image-editing capabilities enabled by our approach. Our code is publicly available at https://github.com/SherryXTChen/TiNO-Edit.
Large Language Models (LLMs) have recently made impressive strides in natural language understanding tasks. Despite their remarkable performance, understanding their decision-making process remains a big challenge. In this paper, we look into bringing some transparency to this process by introducing a new explanation dataset for question answering (QA) tasks that integrates knowledge graphs (KGs) in a novel way. Our dataset includes 12,102 question-answer-explanation (QAE) triples. Each explanation in the dataset links the LLM's reasoning to entities and relations in the KGs. The explanation component includes a why-choose explanation, a why-not-choose explanation, and a set of reason-elements that underlie the LLM's decision. We leverage KGs and graph attention networks (GAT) to find the reason-elements and transform them into why-choose and why-not-choose explanations that are comprehensible to humans. Through quantitative and qualitative evaluations, we demonstrate the potential of our dataset to improve the in-context learning of LLMs, and enhance their interpretability and explainability. Our work contributes to the field of explainable AI by enabling a deeper understanding of the LLMs decision-making process to make them more transparent and thereby, potentially more reliable, to researchers and practitioners alike. Our dataset is available at: https://github.com/chen-zichen/XplainLLM_dataset.git
Large language models (LMs) such as GPT-4 are very powerful and can process different kinds of natural language processing (NLP) tasks. However, it can be difficult to interpret the results due to the multi-layer nonlinear model structure and millions of parameters. Lack of understanding of how the model works can make the model unreliable and dangerous for everyday users in real-world scenarios. Most recent works exploit the weights of attention to provide explanations for model predictions. However, pure attention-based explanation is unable to support the growing complexity of the models, and cannot reason about their decision-making processes. Thus, we propose LMExplainer, a knowledge-enhanced interpretation module for language models that can provide human-understandable explanations. We use a knowledge graph (KG) and a graph attention neural network to extract the key decision signals of the LM. We further explore whether interpretation can also help AI understand the task better. Our experimental results show that LMExplainer outperforms existing LM+KG methods on CommonsenseQA and OpenBookQA. We also compare the explanation results with generated explanation methods and human-annotated results. The comparison shows our method can provide more comprehensive and clearer explanations. LMExplainer demonstrates the potential to enhance model performance and furnish explanations for the reasoning processes of models in natural language.
3D human motion prediction is a research area of high significance and a challenge in computer vision. It is useful for the design of many applications including robotics and autonomous driving. Traditionally, autogregressive models have been used to predict human motion. However, these models have high computation needs and error accumulation that make it difficult to use them for realtime applications. In this paper, we present a non-autogressive model for human motion prediction. We focus on learning spatio-temporal representations non-autoregressively for generation of plausible future motions. We propose a novel architecture that leverages the recently proposed Transformers. Human motion involves complex spatio-temporal dynamics with joints affecting the position and rotation of each other even though they are not connected directly. The proposed model extracts these dynamics using both convolutions and the self-attention mechanism. Using specialized spatial and temporal self-attention to augment the features extracted through convolution allows our model to generate spatio-temporally coherent predictions in parallel independent of the activity. Our contributions are threefold: (i) we frame human motion prediction as a sequence-to-sequence problem and propose a non-autoregressive Transformer to forecast a sequence of poses in parallel; (ii) our method is activity agnostic; (iii) we show that despite its simplicity, our approach is able to make accurate predictions, achieving better or comparable results compared to the state-of-the-art on two public datasets, with far fewer parameters and much faster inference.
Expert-layman text style transfer technologies have the potential to improve communication between members of scientific communities and the general public. High-quality information produced by experts is often filled with difficult jargon laypeople struggle to understand. This is a particularly notable issue in the medical domain, where layman are often confused by medical text online. At present, two bottlenecks interfere with the goal of building high-quality medical expert-layman style transfer systems: a dearth of pretrained medical-domain language models spanning both expert and layman terminologies and a lack of parallel corpora for training the transfer task itself. To mitigate the first issue, we propose a novel language model (LM) pretraining task, Knowledge Base Assimilation, to synthesize pretraining data from the edges of a graph of expert- and layman-style medical terminology terms into an LM during self-supervised learning. To mitigate the second issue, we build a large-scale parallel corpus in the medical expert-layman domain using a margin-based criterion. Our experiments show that transformer-based models pretrained on knowledge base assimilation and other well-established pretraining tasks fine-tuning on our new parallel corpus leads to considerable improvement against expert-layman transfer benchmarks, gaining an average relative improvement of our human evaluation, the Overall Success Rate (OSR), by 106%.
Video represents the majority of internet traffic today leading to a continuous technological arms race between generating higher quality content, transmitting larger file sizes and supporting network infrastructure. Adding to this is the recent COVID-19 pandemic fueled surge in the use of video conferencing tools. Since videos take up substantial bandwidth (~100 Kbps to few Mbps), improved video compression can have a substantial impact on network performance for live and pre-recorded content, providing broader access to multimedia content worldwide. In this work, we present a novel video compression pipeline, called Txt2Vid, which substantially reduces data transmission rates by compressing webcam videos ("talking-head videos") to a text transcript. The text is transmitted and decoded into a realistic reconstruction of the original video using recent advances in deep learning based voice cloning and lip syncing models. Our generative pipeline achieves two to three orders of magnitude reduction in the bitrate as compared to the standard audio-video codecs (encoders-decoders), while maintaining equivalent Quality-of-Experience based on a subjective evaluation by users (n=242) in an online study. The code for this work is available at https://github.com/tpulkit/txt2vid.git.