Interpreting and Analyzing CLIP's Zero-Shot Image Classification via Mutual Knowledge

Contrastive Language-Image Pretraining (CLIP) performs zero-shot image classification by mapping images and textual class representation into a shared embedding space, then retrieving the class closest to the image. This work provides a new approach for interpreting CLIP models for image classification from the lens of mutual knowledge between the two modalities. Specifically, we ask: what concepts do both vision and language CLIP encoders learn in common that influence the joint embedding space, causing points to be closer or further apart? We answer this question via an approach of textual concept-based explanations, showing their effectiveness, and perform an analysis encompassing a pool of 13 CLIP models varying in architecture, size and pretraining datasets. We explore those different aspects in relation to mutual knowledge, and analyze zero-shot predictions. Our approach demonstrates an effective and human-friendly way of understanding zero-shot classification decisions with CLIP.

Learned layered coding for Successive Refinement in the Wyner-Ziv Problem

We propose a data-driven approach to explicitly learn the progressive encoding of a continuous source, which is successively decoded with increasing levels of quality and with the aid of correlated side information. This setup refers to the successive refinement of the Wyner-Ziv coding problem. Assuming ideal Slepian-Wolf coding, our approach employs recurrent neural networks (RNNs) to learn layered encoders and decoders for the quadratic Gaussian case. The models are trained by minimizing a variational bound on the rate-distortion function of the successively refined Wyner-Ziv coding problem. We demonstrate that RNNs can explicitly retrieve layered binning solutions akin to scalable nested quantization. Moreover, the rate-distortion performance of the scheme is on par with the corresponding monolithic Wyner-Ziv coding approach and is close to the rate-distortion bound.

Holistic Representation Learning for Multitask Trajectory Anomaly Detection

Video anomaly detection deals with the recognition of abnormal events in videos. Apart from the visual signal, video anomaly detection has also been addressed with the use of skeleton sequences. We propose a holistic representation of skeleton trajectories to learn expected motions across segments at different times. Our approach uses multitask learning to reconstruct any continuous unobserved temporal segment of the trajectory allowing the extrapolation of past or future segments and the interpolation of in-between segments. We use an end-to-end attention-based encoder-decoder. We encode temporally occluded trajectories, jointly learn latent representations of the occluded segments, and reconstruct trajectories based on expected motions across different temporal segments. Extensive experiments on three trajectory-based video anomaly detection datasets show the advantages and effectiveness of our approach with state-of-the-art results on anomaly detection in skeleton trajectories.

Uni-NLX: Unifying Textual Explanations for Vision and Vision-Language Tasks

Natural Language Explanations (NLE) aim at supplementing the prediction of a model with human-friendly natural text. Existing NLE approaches involve training separate models for each downstream task. In this work, we propose Uni-NLX, a unified framework that consolidates all NLE tasks into a single and compact multi-task model using a unified training objective of text generation. Additionally, we introduce two new NLE datasets: 1) ImageNetX, a dataset of 144K samples for explaining ImageNet categories, and 2) VQA-ParaX, a dataset of 123K samples for explaining the task of Visual Question Answering (VQA). Both datasets are derived leveraging large language models (LLMs). By training on the 1M combined NLE samples, our single unified framework is capable of simultaneously performing seven NLE tasks including VQA, visual recognition and visual reasoning tasks with 7X fewer parameters, demonstrating comparable performance to the independent task-specific models in previous approaches, and in certain tasks even outperforming them. Code is at https://github.com/fawazsammani/uni-nlx

The STOIC2021 COVID-19 AI challenge: applying reusable training methodologies to private data

Challenges drive the state-of-the-art of automated medical image analysis. The quantity of public training data that they provide can limit the performance of their solutions. Public access to the training methodology for these solutions remains absent. This study implements the Type Three (T3) challenge format, which allows for training solutions on private data and guarantees reusable training methodologies. With T3, challenge organizers train a codebase provided by the participants on sequestered training data. T3 was implemented in the STOIC2021 challenge, with the goal of predicting from a computed tomography (CT) scan whether subjects had a severe COVID-19 infection, defined as intubation or death within one month. STOIC2021 consisted of a Qualification phase, where participants developed challenge solutions using 2000 publicly available CT scans, and a Final phase, where participants submitted their training methodologies with which solutions were trained on CT scans of 9724 subjects. The organizers successfully trained six of the eight Final phase submissions. The submitted codebases for training and running inference were released publicly. The winning solution obtained an area under the receiver operating characteristic curve for discerning between severe and non-severe COVID-19 of 0.815. The Final phase solutions of all finalists improved upon their Qualification phase solutions.HSUXJM-TNZF9CHSUXJM-TNZF9C

Leaping Into Memories: Space-Time Deep Feature Synthesis

The success of deep learning models has led to their adaptation and adoption by prominent video understanding methods. The majority of these approaches encode features in a joint space-time modality for which the inner workings and learned representations are difficult to visually interpret. We propose LEArned Preconscious Synthesis (LEAPS), an architecture-agnostic method for synthesizing videos from the internal spatiotemporal representations of models. Using a stimulus video and a target class, we prime a fixed space-time model and iteratively optimize a video initialized with random noise. We incorporate additional regularizers to improve the feature diversity of the synthesized videos as well as the cross-frame temporal coherence of motions. We quantitatively and qualitatively evaluate the applicability of LEAPS by inverting a range of spatiotemporal convolutional and attention-based architectures trained on Kinetics-400, which to the best of our knowledge has not been previously accomplished.

Entropy-Based Feature Extraction For Real-Time Semantic Segmentation

This paper introduces an efficient patch-based computational module, coined Entropy-based Patch Encoder (EPE) module, for resource-constrained semantic segmentation. The EPE module consists of three lightweight fully-convolutional encoders, each extracting features from image patches with a different amount of entropy. Patches with high entropy are being processed by the encoder with the largest number of parameters, patches with moderate entropy are processed by the encoder with a moderate number of parameters, and patches with low entropy are processed by the smallest encoder. The intuition behind the module is the following: as patches with high entropy contain more information, they need an encoder with more parameters, unlike low entropy patches, which can be processed using a small encoder. Consequently, processing part of the patches via the smaller encoder can significantly reduce the computational cost of the module. Experiments show that EPE can boost the performance of existing real-time semantic segmentation models with a slight increase in the computational cost. Specifically, EPE increases the mIOU performance of DFANet A by 0.9% with only 1.2% increase in the number of parameters and the mIOU performance of EDANet by 1% with 10% increase of the model parameters.

Representation Learning with Information Theory for COVID-19 Detection

Successful data representation is a fundamental factor in machine learning based medical imaging analysis. Deep Learning (DL) has taken an essential role in robust representation learning. However, the inability of deep models to generalize to unseen data can quickly overfit intricate patterns. Thereby, we can conveniently implement strategies to aid deep models in discovering useful priors from data to learn their intrinsic properties. Our model, which we call a dual role network (DRN), uses a dependency maximization approach based on Least Squared Mutual Information (LSMI). The LSMI leverages dependency measures to ensure representation invariance and local smoothness. While prior works have used information theory measures like mutual information, known to be computationally expensive due to a density estimation step, our LSMI formulation alleviates the issues of intractable mutual information estimation and can be used to approximate it. Experiments on CT based COVID-19 Detection and COVID-19 Severity Detection benchmarks demonstrate the effectiveness of our method.

Visualizing and Understanding Self-Supervised Vision Learning

Self-Supervised vision learning has revolutionized deep learning, becoming the next big challenge in the domain and rapidly closing the gap with supervised methods on large computer vision benchmarks. With current models and training data exponentially growing, explaining and understanding these models becomes pivotal. We study the problem of explainable artificial intelligence in the domain of self-supervised learning for vision tasks, and present methods to understand networks trained with self-supervision and their inner workings. Given the huge diversity of self-supervised vision pretext tasks, we narrow our focus on understanding paradigms which learn from two views of the same image, and mainly aim to understand the pretext task. Our work focuses on explaining similarity learning, and is easily extendable to all other pretext tasks. We study two popular self-supervised vision models: SimCLR and Barlow Twins. We develop a total of six methods for visualizing and understanding these models: Perturbation-based methods (conditional occlusion, context-agnostic conditional occlusion and pairwise occlusion), Interaction-CAM, Feature Visualization, Model Difference Visualization, Averaged Transforms and Pixel Invaraince. Finally, we evaluate these explanations by translating well-known evaluation metrics tailored towards supervised image classification systems involving a single image, into the domain of self-supervised learning where two images are involved. Code is at: https://github.com/fawazsammani/xai-ssl

NLX-GPT: A Model for Natural Language Explanations in Vision and Vision-Language Tasks

Natural language explanation (NLE) models aim at explaining the decision-making process of a black box system via generating natural language sentences which are human-friendly, high-level and fine-grained. Current NLE models explain the decision-making process of a vision or vision-language model (a.k.a., task model), e.g., a VQA model, via a language model (a.k.a., explanation model), e.g., GPT. Other than the additional memory resources and inference time required by the task model, the task and explanation models are completely independent, which disassociates the explanation from the reasoning process made to predict the answer. We introduce NLX-GPT, a general, compact and faithful language model that can simultaneously predict an answer and explain it. We first conduct pre-training on large scale data of image-caption pairs for general understanding of images, and then formulate the answer as a text prediction task along with the explanation. Without region proposals nor a task model, our resulting overall framework attains better evaluation scores, contains much less parameters and is 15$\times$ faster than the current SoA model. We then address the problem of evaluating the explanations which can be in many times generic, data-biased and can come in several forms. We therefore design 2 new evaluation measures: (1) explain-predict and (2) retrieval-based attack, a self-evaluation framework that requires no labels. Code is at: https://github.com/fawazsammani/nlxgpt.