Uncertainty-Aware Multi-Shot Knowledge Distillation for Image-Based Object Re-Identification

Object re-identification (re-id) aims to identify a specific object across times or camera views, with the person re-id and vehicle re-id as the most widely studied applications. Re-id is challenging because of the variations in viewpoints, (human) poses, and occlusions. Multi-shots of the same object can cover diverse viewpoints/poses and thus provide more comprehensive information. In this paper, we propose exploiting the multi-shots of the same identity to guide the feature learning of each individual image. Specifically, we design an Uncertainty-aware Multi-shot Teacher-Student (UMTS) Network. It consists of a teacher network (T-net) that learns the comprehensive features from multiple images of the same object, and a student network (S-net) that takes a single image as input. In particular, we take into account the data dependent heteroscedastic uncertainty for effectively transferring the knowledge from the T-net to S-net. To the best of our knowledge, we are the first to make use of multi-shots of an object in a teacher-student learning manner for effectively boosting the single image based re-id. We validate the effectiveness of our approach on the popular vehicle re-id and person re-id datasets. In inference, the S-net alone significantly outperforms the baselines and achieves the state-of-the-art performance.

Download Full-text

Collisions and Damage of Offshore Structures: A State-of-the-Art

Journal of Energy Resources Technology ◽

10.1115/1.3231193 ◽

1985 ◽

Vol 107 (3) ◽

pp. 297-314 ◽

Cited By ~ 19

Author(s):

C. P. Ellinas ◽

S. Valsgard

Keyword(s):

State Of The Art ◽

Offshore Structures ◽

The State ◽

Design Parameters ◽

Future Research ◽

Research Activity ◽

Comprehensive Information ◽

Major Interest ◽

Offshore Industry ◽

Available Information

Over the recent years, following the very rapid increase in the construction and installation of offshore structures, there has been a considerable growth of interest in the assessment of the probabilities and consequences of collision and damage of such structures. This is reflected by the very large number of papers published over the last 15 yr and the multitude of conferences and meetings held on the subject. Many research programs have been completed or are in progress at many centers and institutions over the world. Accidental loading and damage are now accepted design parameters recommended for consideration in a number of Codes for the design in offshore structures. This paper reviews the state-of-the-art with respect to the probabilities and consequences of collisions and accidental loading in general, and methods for the assessment of the design of steel offshore structures against damage. Most of the available information in the field of offshore collisions and accidental loading emanates from research and experience related to ship safety. However, in this paper emphasis is placed on research activity and available information concerned with offshore structures, such as platforms, semisubmersibles, etc. There is a considerable amount of information available on methods for evaluating the extent and effects on damage of these structures and in estimating their residual strength in the damaged condition. As this is an area currently of major interest in the offshore industry, the paper presents comprehensive information and some new results relating to all major structural components. The state-of-the-art with regards to methods and principles for design against damage is also reviewed and commented upon. The paper concludes with general recommendations and indications of areas where future research could be most usefully directed.

Download Full-text

Informative Multimodal Unsupervised Image-to-Image Translation

10.5121/csit.2021.110503 ◽

2021 ◽

Author(s):

Tien Tai Doan ◽

Guillaume Ghyselinck ◽

Blaise Hanczar

Keyword(s):

Image Quality ◽

Mutual Information ◽

State Of The Art ◽

The State ◽

New Method ◽

Target Domain ◽

Multiple Images ◽

Source Domain ◽

Multiple Image ◽

Image Translation

We propose a new method of multimodal image translation, called InfoMUNIT, which is an extension of the state-of-the-art method MUNIT. Our method allows controlling the style of the generated images and improves their quality and diversity. It learns to maximize the mutual information between a subset of style code and the distribution of the output images. Experiments show that our model cannot only translate one image from the source domain to multiple images in the target domain but also explore and manipulate features of the outputs without annotation. Furthermore, it achieves a superior diversity and a competitive image quality to state-of-the-art methods in multiple image translation tasks.

Download Full-text

Single Image Defogging Method Based on Image Patch Decomposition and Multi-Exposure Image Fusion

Frontiers in Neurorobotics ◽

10.3389/fnbot.2021.700483 ◽

2021 ◽

Vol 15 ◽

Author(s):

Qiuzhuo Liu ◽

Yaqin Luo ◽

Ke Li ◽

Wenfeng Li ◽

Yi Chai ◽

...

Keyword(s):

Image Restoration ◽

Physical Model ◽

State Of The Art ◽

The State ◽

Depth Information ◽

Single Image ◽

Image Patch ◽

Image Defogging ◽

Image Patches ◽

Scene Depth

Bad weather conditions (such as fog, haze) seriously affect the visual quality of images. According to the scene depth information, physical model-based methods are used to improve image visibility for further image restoration. However, the unstable acquisition of the scene depth information seriously affects the defogging performance of physical model-based methods. Additionally, most of image enhancement-based methods focus on the global adjustment of image contrast and saturation, and lack the local details for image restoration. So, this paper proposes a single image defogging method based on image patch decomposition and multi-exposure fusion. First, a single foggy image is processed by gamma correction to obtain a set of underexposed images. Then the saturation of the obtained underexposed and original images is enhanced. Next, each image in the multi-exposure image set (including the set of underexposed images and the original image) is decomposed into the base and detail layers by a guided filter. The base layers are first decomposed into image patches, and then the fusion weight maps of the image patches are constructed. For detail layers, the exposure features are first extracted from the luminance components of images, and then the extracted exposure features are evaluated by constructing gaussian functions. Finally, both base and detail layers are combined to obtain the defogged image. The proposed method is compared with the state-of-the-art methods. The comparative experimental results confirm the effectiveness of the proposed method and its superiority over the state-of-the-art methods.

Download Full-text

An Adaptive Framework for Conversational Question Answering

Proceedings of the AAAI Conference on Artificial Intelligence ◽

10.1609/aaai.v33i01.330110041 ◽

2019 ◽

Vol 33 ◽

pp. 10041-10042

Author(s):

Lixin Su ◽

Jiafeng Guo ◽

Yixing Fan ◽

Yanyan Lan ◽

Ruqing Zhang ◽

...

Keyword(s):

Preliminary Analysis ◽

Information Needs ◽

Question Answering ◽

State Of The Art ◽

The State ◽

Dominant Type ◽

New Knowledge ◽

Specific Object ◽

Overall Performance ◽

Adaptive Framework

In Conversational Question Answering (CoQA), humans propose a series of questions to satisfy their information needs. Based on our preliminary analysis, there are two major types of questions, namely verification questions and knowledgeseeking questions. The first one is to verify some existing facts, while the latter is to obtain new knowledge about some specific object. These two types of questions differ significantly in their answering ways. However, existing methods usually treat them uniformly, which may easily be biased by the dominant type of questions and obtain inferior overall performance. In this work, we propose an adaptive framework to handle these two types of questions in different ways based on their own characteristics. We conduct experiments on the recently released CoQA benchmark dataset, and the results demonstrate that our method outperforms the state-of-the-art baseline methods.

Download Full-text

Hierarchical Knowledge Squeezed Adversarial Network Compression

Proceedings of the AAAI Conference on Artificial Intelligence ◽

10.1609/aaai.v34i07.6799 ◽

2020 ◽

Vol 34 (07) ◽

pp. 11370-11377

Author(s):

Peng Li ◽

Chang Shu ◽

Yuan Xie ◽

Yan Qu ◽

Hui Kong

Keyword(s):

State Of The Art ◽

Teacher Student ◽

Adversarial Network ◽

Benchmark Datasets ◽

Knowledge Distillation ◽

Adversarial Training ◽

Rich Information ◽

Process Oriented ◽

Transfer Method ◽

Network Compression

Deep network compression has been achieved notable progress via knowledge distillation, where a teacher-student learning manner is adopted by using predetermined loss. Recently, more focuses have been transferred to employ the adversarial training to minimize the discrepancy between distributions of output from two networks. However, they always emphasize on result-oriented learning while neglecting the scheme of process-oriented learning, leading to the loss of rich information contained in the whole network pipeline. Whereas in other (non GAN-based) process-oriented methods, the knowledge have usually been transferred in a redundant manner. Observing that, the small network can not perfectly mimic a large one due to the huge gap of network scale, we propose a knowledge transfer method, involving effective intermediate supervision, under the adversarial training framework to learn the student network. Different from the other intermediate supervision methods, we design the knowledge representation in a compact form by introducing a task-driven attention mechanism. Meanwhile, to improve the representation capability of the attention-based method, a hierarchical structure is utilized so that powerful but highly squeezed knowledge is realized and the knowledge from teacher network could accommodate the size of student network. Extensive experimental results on three typical benchmark datasets, i.e., CIFAR-10, CIFAR-100, and ImageNet, demonstrate that our method achieves highly superior performances against state-of-the-art methods.

Download Full-text

MiDTD: A Simple and Effective Distillation Framework for Distantly Supervised Relation Extraction

ACM Transactions on Information Systems ◽

10.1145/3503917 ◽

2022 ◽

Vol 40 (4) ◽

pp. 1-32

Author(s):

Rui Li ◽

Cheng Yang ◽

Tingwei Li ◽

Sen Su

Keyword(s):

Information Extraction ◽

Temperature Regulation ◽

State Of The Art ◽

Relation Extraction ◽

Dynamic Temperature ◽

Teacher Student ◽

Distant Supervision ◽

Annotation Data ◽

Knowledge Distillation ◽

Moderate Range

Relation extraction (RE), an important information extraction task, faced the great challenge brought by limited annotation data. To this end, distant supervision was proposed to automatically label RE data, and thus largely increased the number of annotated instances. Unfortunately, lots of noise relation annotations brought by automatic labeling become a new obstacle. Some recent studies have shown that the teacher-student framework of knowledge distillation can alleviate the interference of noise relation annotations via label softening. Nevertheless, we find that they still suffer from two problems: propagation of inaccurate dark knowledge and constraint of a unified distillation temperature . In this article, we propose a simple and effective Multi-instance Dynamic Temperature Distillation (MiDTD) framework, which is model-agnostic and mainly involves two modules: multi-instance target fusion (MiTF) and dynamic temperature regulation (DTR). MiTF combines the teacher’s predictions for multiple sentences with the same entity pair to amend the inaccurate dark knowledge in each student’s target. DTR allocates alterable distillation temperatures to different training instances to enable the softness of most student’s targets to be regulated to a moderate range. In experiments, we construct three concrete MiDTD instantiations with BERT, PCNN, and BiLSTM-based RE models, and the distilled students significantly outperform their teachers and the state-of-the-art (SOTA) methods.

Download Full-text

Progressive Blockwise Knowledge Distillation for Neural Network Acceleration

Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence ◽

10.24963/ijcai.2018/384 ◽

2018 ◽

Cited By ~ 5

Author(s):

Hui Wang ◽

Hanbin Zhao ◽

Xi Li ◽

Xu Tan

Keyword(s):

Neural Network ◽

Function Approximation ◽

State Of The Art ◽

Design Criterion ◽

Structure Design ◽

Model Accuracy ◽

Teacher Student ◽

Block Level ◽

Knowledge Distillation ◽

Teacher Network

As an important and challenging problem in machine learning and computer vision, neural network acceleration essentially aims to enhance the computational efficiency without sacrificing the model accuracy too much. In this paper, we propose a progressive blockwise learning scheme for teacher-student model distillation at the subnetwork block level. The proposed scheme is able to distill the knowledge of the entire teacher network by locally extracting the knowledge of each block in terms of progressive blockwise function approximation. Furthermore, we propose a structure design criterion for the student subnetwork block, which is able to effectively preserve the original receptive field from the teacher network. Experimental results demonstrate the effectiveness of the proposed scheme against the state-of-the-art approaches.

Download Full-text

Practical Picture Processing

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100051700 ◽

1974 ◽

Vol 32 ◽

pp. 338-339

Author(s):

T. A. Welton

Keyword(s):

Radiation Damage ◽

Coherence Length ◽

Spatial Information ◽

State Of The Art ◽

Coherent Radiation ◽

The State ◽

Energy Spread ◽

Electron Micrograph ◽

Picture Processing ◽

Molecular Skeleton

Various authors have emphasized the spatial information resident in an electron micrograph taken with adequately coherent radiation. In view of the completion of at least one such instrument, this opportunity is taken to summarize the state of the art of processing such micrographs. We use the usual symbols for the aberration coefficients, and supplement these with £ and 6 for the transverse coherence length and the fractional energy spread respectively. He also assume a weak, biologically interesting sample, with principal interest lying in the molecular skeleton remaining after obvious hydrogen loss and other radiation damage has occurred.

Download Full-text