arXiv subjects

Li Fei-Fei

Publications and source records attributed to Li Fei-Fei.

At least 235 records · Page 13Linked to original sources

arXiv2016

Towards Viewpoint Invariant 3D Human Pose Estimation

We propose a viewpoint invariant model for 3D human pose estimation from a single depth image. To achieve this, our discriminative model embeds local regions into a learned viewpoint invariant feature space. Formulated as a multi-task learning problem, our model is able to selectively predict partial poses in the presence of noise and occlusion. Our approach leverages a convolutional and recurrent network architecture with a top-down error feedback mechanism to self-correct previous pose estimates in an end-to-end manner. We evaluate our model on a previously published depth dataset and a newly collected human pose dataset containing 100K annotated depth images from extreme viewpoints. Experiments show that our model achieves competitive performance on frontal views while achieving state-of-the-art performance on alternate viewpoints.

cs.CV↗

arXiv2016

What's the Point: Semantic Segmentation with Point Supervision

The semantic image segmentation task presents a trade-off between test time accuracy and training-time annotation cost. Detailed per-pixel annotations enable training accurate models but are very time-consuming to obtain, image-level class labels are an order of magnitude cheaper but result in less accurate models. We take a natural step from image-level annotation towards stronger supervision: we ask annotators to point to an object if one exists. We incorporate this point supervision along with a novel objectness potential in the training loss function of a CNN model. Experimental results on the PASCAL VOC 2012 benchmark reveal that the combined effect of point-level supervision and objectness potential yields an improvement of 12.9% mIOU over image-level supervision. Further, we demonstrate that models trained with point-level supervision are more accurate than models trained with image-level, squiggle-level or full supervision given a fixed annotation budget.

cs.CV↗

arXiv2016

Locally-Optimized Inter-Subject Alignment of Functional Cortical Regions

Inter-subject registration of cortical areas is necessary in functional imaging (fMRI) studies for making inferences about equivalent brain function across a population. However, many high-level visual brain areas are defined as peaks of functional contrasts whose cortical position is highly variable. As such, most alignment methods fail to accurately map functional regions of interest (ROIs) across participants. To address this problem, we propose a locally optimized registration method that directly predicts the location of a seed ROI on a separate target cortical sheet by maximizing the functional correlation between their time courses, while simultaneously allowing for non-smooth local deformations in region topology. Our method outperforms the two most commonly used alternatives (anatomical landmark-based AFNI alignment and cortical convexity-based FreeSurfer alignment) in overlap between predicted region and functionally-defined LOC. Furthermore, the maps obtained using our method are more consistent across subjects than both baseline measures. Critically, our method represents an important step forward towards predicting brain regions without explicit localizer scans and deciphering the poorly understood relationship between the location of functional regions, their anatomical extent, and the consistency of computations those regions perform across people.

q-bio.NC↗

arXiv2016

Visual7W: Grounded Question Answering in Images

We have seen great progress in basic perceptual tasks such as object recognition and detection. However, AI models still fail to match humans in high-level vision tasks due to the lack of capacities for deeper reasoning. Recently the new task of visual question answering (QA) has been proposed to evaluate a model's capacity for deep image understanding. Previous works have established a loose, global association between QA sentences and images. However, many questions and answers, in practice, relate to local regions in the images. We establish a semantic link between textual descriptions and image regions by object-level grounding. It enables a new type of QA with visual answers, in addition to textual answers used in previous work. We study the visual QA tasks in a grounded setting with a large collection of 7W multiple-choice QA pairs. Furthermore, we evaluate human performance and several baseline models on the QA tasks. Finally, we propose a novel LSTM model with spatial attention to tackle the 7W QA tasks.

cs.CV↗

arXiv2016

Perceptual Losses for Real-Time Style Transfer and Super-Resolution

We consider image transformation problems, where an input image is transformed into an output image. Recent methods for such problems typically train feed-forward convolutional neural networks using a \emph{per-pixel} loss between the output and ground-truth images. Parallel work has shown that high-quality images can be generated by defining and optimizing \emph{perceptual} loss functions based on high-level features extracted from pretrained networks. We combine the benefits of both approaches, and propose the use of perceptual loss functions for training feed-forward networks for image transformation tasks. We show results on image style transfer, where a feed-forward network is trained to solve the optimization problem proposed by Gatys et al in real-time. Compared to the optimization-based method, our network gives similar qualitative results but is three orders of magnitude faster. We also experiment with single-image super-resolution, where replacing a per-pixel loss with a perceptual loss gives visually pleasing results.

cs.CV↗

arXiv2016

Detecting events and key actors in multi-person videos

Multi-person event recognition is a challenging task, often with many people active in the scene but only a small subset contributing to an actual event. In this paper, we propose a model which learns to detect events in such videos while automatically "attending" to the people responsible for the event. Our model does not use explicit annotations regarding who or where those people are during training and testing. In particular, we track people in videos and use a recurrent neural network (RNN) to represent the track features. We learn time-varying attention weights to combine these features at each time-instant. The attended features are then processed using another RNN for event detection/classification. Since most video datasets with multiple people are restricted to a small number of videos, we also collected a new basketball dataset comprising 257 basketball games with 14K event annotations corresponding to 11 event classes. Our model outperforms state-of-the-art methods for both event classification and detection on this new dataset. Additionally, we show that the attention mechanism is able to consistently localize the relevant players.

cs.CV↗

arXiv2016

Embracing Error to Enable Rapid Crowdsourcing

Microtask crowdsourcing has enabled dataset advances in social science and machine learning, but existing crowdsourcing schemes are too expensive to scale up with the expanding volume of data. To scale and widen the applicability of crowdsourcing, we present a technique that produces extremely rapid judgments for binary and categorical labels. Rather than punishing all errors, which causes workers to proceed slowly and deliberately, our technique speeds up workers' judgments to the point where errors are acceptable and even expected. We demonstrate that it is possible to rectify these errors by randomizing task order and modeling response latency. We evaluate our technique on a breadth of common labeling tasks such as image verification, word similarity, sentiment analysis and topic classification. Where prior work typically achieves a 0.25x to 1x speedup over fixed majority vote, our approach often achieves an order of magnitude (10x) speedup.

cs.HC↗

arXiv2015

DenseCap: Fully Convolutional Localization Networks for Dense Captioning

We introduce the dense captioning task, which requires a computer vision system to both localize and describe salient regions in images in natural language. The dense captioning task generalizes object detection when the descriptions consist of a single word, and Image Captioning when one predicted region covers the full image. To address the localization and description task jointly we propose a Fully Convolutional Localization Network (FCLN) architecture that processes an image with a single, efficient forward pass, requires no external regions proposals, and can be trained end-to-end with a single round of optimization. The architecture is composed of a Convolutional Network, a novel dense localization layer, and Recurrent Neural Network language model that generates the label sequences. We evaluate our network on the Visual Genome dataset, which comprises 94,000 images and 4,100,000 region-grounded captions. We observe both speed and accuracy improvements over baselines based on current state of the art approaches in both generation and retrieval settings.

cs.CV↗

arXiv2015

Visualizing and Understanding Recurrent Networks

Recurrent Neural Networks (RNNs), and specifically a variant with Long Short-Term Memory (LSTM), are enjoying renewed interest as a result of successful applications in a wide range of machine learning problems that involve sequential data. However, while LSTMs provide exceptional results in practice, the source of their performance and their limitations remain rather poorly understood. Using character-level language models as an interpretable testbed, we aim to bridge this gap by providing an analysis of their representations, predictions and error types. In particular, our experiments reveal the existence of interpretable cells that keep track of long-range dependencies such as line lengths, quotes and brackets. Moreover, our comparative analysis with finite horizon n-gram models traces the source of the LSTM improvements to long-range structural dependencies. Finally, we provide analysis of the remaining errors and suggests areas for further study.

cs.LG↗

arXiv2015

Building a Large-scale Multimodal Knowledge Base System for Answering Visual Queries

The complexity of the visual world creates significant challenges for comprehensive visual understanding. In spite of recent successes in visual recognition, today's vision systems would still struggle to deal with visual queries that require a deeper reasoning. We propose a knowledge base (KB) framework to handle an assortment of visual queries, without the need to train new classifiers for new tasks. Building such a large-scale multimodal KB presents a major challenge of scalability. We cast a large-scale MRF into a KB representation, incorporating visual, textual and structured data, as well as their diverse relations. We introduce a scalable knowledge base construction system that is capable of building a KB with half billion variables and millions of parameters in a few hours. Our system achieves competitive results compared to purpose-built models on standard recognition and retrieval tasks, while exhibiting greater flexibility in answering richer visual queries.

cs.CV↗

arXiv2015

SentenceRacer: A Game with a Purpose for Image Sentence Annotation

Recently datasets that contain sentence descriptions of images have enabled models that can automatically generate image captions. However, collecting these datasets are still very expensive. Here, we present SentenceRacer, an online game that gathers and verifies descriptions of images at no cost. Similar to the game hangman, players compete to uncover words in a sentence that ultimately describes an image. SentenceRacer both generates and verifies that the sentences are accurate descriptions. We show that SentenceRacer generates annotations of higher quality than those generated on Amazon Mechanical Turk (AMT).

cs.HC↗

arXiv2015

Improving Image Classification with Location Context

With the widespread availability of cellphones and cameras that have GPS capabilities, it is common for images being uploaded to the Internet today to have GPS coordinates associated with them. In addition to research that tries to predict GPS coordinates from visual features, this also opens up the door to problems that are conditioned on the availability of GPS coordinates. In this work, we tackle the problem of performing image classification with location context, in which we are given the GPS coordinates for images in both the train and test phases. We explore different ways of encoding and extracting features from the GPS coordinates, and show how to naturally incorporate these features into a Convolutional Neural Network (CNN), the current state-of-the-art for most image classification and recognition problems. We also show how it is possible to simultaneously learn the optimal pooling radii for a subset of our features within the CNN framework. To evaluate our model and to help promote research in this area, we identify a set of location-sensitive concepts and annotate a subset of the Yahoo Flickr Creative Commons 100M dataset that has GPS coordinates with these concepts, which we make publicly available. By leveraging location context, we are able to achieve almost a 7% gain in mean average precision.

cs.CV↗

arXiv2015

Learning Temporal Embeddings for Complex Video Analysis

In this paper, we propose to learn temporal embeddings of video frames for complex video analysis. Large quantities of unlabeled video data can be easily obtained from the Internet. These videos possess the implicit weak label that they are sequences of temporally and semantically coherent images. We leverage this information to learn temporal embeddings for video frames by associating frames with the temporal context that they appear in. To do this, we propose a scheme for incorporating temporal context based on past and future frames in videos, and compare this to other contextual representations. In addition, we show how data augmentation using multi-resolution samples and hard negatives helps to significantly improve the quality of the learned embeddings. We evaluate various design decisions for learning temporal embeddings, and show that our embeddings can improve performance for multiple video tasks such as retrieval, classification, and temporal order recovery in unconstrained Internet video.

cs.CV↗

arXiv2015

Deep Visual-Semantic Alignments for Generating Image Descriptions

We present a model that generates natural language descriptions of images and their regions. Our approach leverages datasets of images and their sentence descriptions to learn about the inter-modal correspondences between language and visual data. Our alignment model is based on a novel combination of Convolutional Neural Networks over image regions, bidirectional Recurrent Neural Networks over sentences, and a structured objective that aligns the two modalities through a multimodal embedding. We then describe a Multimodal Recurrent Neural Network architecture that uses the inferred alignments to learn to generate novel descriptions of image regions. We demonstrate that our alignment model produces state of the art results in retrieval experiments on Flickr8K, Flickr30K and MSCOCO datasets. We then show that the generated descriptions significantly outperform retrieval baselines on both full images and on a new dataset of region-level annotations.

cs.CV↗

arXiv2015

ImageNet Large Scale Visual Recognition Challenge

The ImageNet Large Scale Visual Recognition Challenge is a benchmark in object category classification and detection on hundreds of object categories and millions of images. The challenge has been run annually from 2010 to present, attracting participation from more than fifty institutions. This paper describes the creation of this benchmark dataset and the advances in object recognition that have been possible as a result. We discuss the challenges of collecting large-scale ground truth annotation, highlight key breakthroughs in categorical object recognition, provide a detailed analysis of the current state of the field of large-scale image classification and object detection, and compare the state-of-the-art computer vision accuracy with human accuracy. We conclude with lessons learned in the five years of the challenge, and propose future directions and improvements.

cs.CV↗

arXiv2014

Visual Noise from Natural Scene Statistics Reveals Human Scene Category Representations

Our perceptions are guided both by the bottom-up information entering our eyes, as well as our top-down expectations of what we will see. Although bottom-up visual processing has been extensively studied, comparatively little is known about top-down signals. Here, we describe REVEAL (Representations Envisioned Via Evolutionary ALgorithm), a method for visualizing an observer's internal representation of a complex, real-world scene, allowing us to, for the first time, visualize the top-down information in an observer's mind. REVEAL rests on two innovations for solving this high dimensional problem: visual noise that samples from natural image statistics, and a computer algorithm that collaborates with human observers to efficiently obtain a solution. In this work, we visualize observers' internal representations of a visual scene category (street) using an experiment in which the observer views the naturalistic visual noise and collaborates with the algorithm to externalize his internal representation. As no scene information was presented, observers had to use their internal knowledge of the target, matching it with the visual features in the noise. We matched reconstructed images with images of real-world street scenes to enhance visualization. Critically, we show that the visualized mental images can be used to predict rapid scene detection performance, as each observer had faster and more accurate responses to detecting real-world images that were the most similar to his reconstructed street templates. These results show that it is possible to visualize previously unobservable mental representations of real world stimuli. More broadly, REVEAL provides a general method for objectively examining the content of previously private, subjective mental experiences.

cs.CV↗

arXiv2014

Affordances Provide a Fundamental Categorization Principle for Visual Scenes

How do we know that a kitchen is a kitchen by looking? Relatively little is known about how we conceptualize and categorize different visual environments. Traditional models of visual perception posit that scene categorization is achieved through the recognition of a scene's objects, yet these models cannot account for the mounting evidence that human observers are relatively insensitive to the local details in an image. Psychologists have long theorized that the affordances, or actionable possibilities of a stimulus are pivotal to its perception. To what extent are scene categories created from similar affordances? Using a large-scale experiment using hundreds of scene categories, we show that the activities afforded by a visual scene provide a fundamental categorization principle. Affordance-based similarity explained the majority of the structure in the human scene categorization patterns, outperforming alternative similarities based on objects or visual features. We all models were combined, affordances provided the majority of the predictive power in the combined model, and nearly half of the total explained variance is captured only by affordances. These results challenge many existing models of high-level visual perception, and provide immediately testable hypotheses for the functional organization of the human perceptual system.

q-bio.NC↗

arXiv2014

VideoSET: Video Summary Evaluation through Text

In this paper we present VideoSET, a method for Video Summary Evaluation through Text that can evaluate how well a video summary is able to retain the semantic information contained in its original video. We observe that semantics is most easily expressed in words, and develop a text-based approach for the evaluation. Given a video summary, a text representation of the video summary is first generated, and an NLP-based metric is then used to measure its semantic distance to ground-truth text summaries written by humans. We show that our technique has higher agreement with human judgment than pixel-based distance metrics. We also release text annotations and ground-truth text summaries for a number of publicly available video datasets, for use by the computer vision community.

cs.CV↗