Filters








16,821 Hits in 5.7 sec

Detecting Twenty-thousand Classes using Image-level Supervision [article]

Xingyi Zhou, Rohit Girdhar, Armand Joulin, Philipp Krähenbühl, Ishan Misra
2022 arXiv   pre-print
For the first time, we train a detector with all the twenty-one-thousand classes of the ImageNet dataset and show that it generalizes to new datasets without finetuning. Code is available at .  ...  We propose Detic, which simply trains the classifiers of a detector on image classification data and thus expands the vocabulary of detectors to tens of thousands of concepts.  ...  Finally, we train a detector using the full ImageNet-21K with more than twenty-thousand classes.  ... 
arXiv:2201.02605v3 fatcat:fsxcztxd2rdthoxgxjwaz4gbxe

Detecting Twenty-thousand Classes using Image-level Supervision [article]

Xingyi Zhou, Rohit Girdha, Armand Joulin, Phillip Krähenbühl, Ishan Misra
2022
For the first time, we train a detector with all the twenty-one-thousand classes of the ImageNet dataset and show that it generalizes to new datasets without fine-tuning.  ...  We propose Detic, which simply trains the classifiers of a detector on image classification data and thus expands the vocabulary of detectors to tens of thousands of concepts.  ...  Finally, we train a detector using the full ImageNet-21K with more than twenty-thousand classes.  ... 
doi:10.48550/arxiv.2201.02605 fatcat:bftwc7qeejevpgv6dtvilul7wq

A Quantitative Analysis Platform for PD-L1 Immunohistochemistry based on Point-level Supervision Model

Haibo Mi, Kele Xu, Yang Xiang, Yulin He, Dawei Feng, Huaimin Wang, Chun Wu, Yanming Song, Xiaolei Sun
2019 Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence  
As point-level annotations can provide a rough estimate of the object locations and classifications, this platform adopts a point-level supervision model to classify, localize, and count the PD-L1 cells  ...  In this paper, we describe the development of a platform for PD-L1 pathological image quantitative analysis using deep learning approaches.  ...  We define an image-level loss to classify the nuclei into the negative ones and positive ones.  ... 
doi:10.24963/ijcai.2019/954 dblp:conf/ijcai/MiXXHFWWSS19 fatcat:whtpdl4ex5eanhqivwital5gou

Natural Disaster Classification using Aerial Photography Explainable for Typhoon Damaged Feature [article]

Takato Yasuno, Masazumi Amakata, Masahiro Okano
2020 arXiv   pre-print
Furthermore, we can realize explainable map on each unit grid images to compute the convolutional activation map using Grad-CAM.  ...  Using target feature class probabilities, we can visualize disaster feature map to scale a color range.  ...  We thank Takuji Fukumoto and Shinichi Kuramoto for supporting us MATLAB resources.  ... 
arXiv:2004.10130v5 fatcat:jdgo62c33jhuzfadmqvkkxiqqe

Towards to Reasonable Decision Basis in Automatic Bone X-Ray Image Classification: A Weakly-Supervised Approach

Jianjie Lu, Kai-yu Tong
2019 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
A weakly-supervised framework is proposed that cannot only make class inference but also provides reasonable decision basis in bone X-ray images.  ...  We implement it in three stages progressively: (1) design a classification network and use positive class activation map (PCAM) for attention location; (2) generate masks from attention maps and lead the  ...  MURA Dataset: Towards Radiologist-level Abnormality Detection in Musculoskeletal Radiographs. In International Conference on Medical Imaging with Deep Learning.  ... 
doi:10.1609/aaai.v33i01.33019985 fatcat:2qzhqekilvfdll4nyz2xzso2qm

AnoNet: Weakly Supervised Anomaly Detection in Textured Surfaces [article]

Manpreet Singh Minhas, John Zelek
2019 arXiv   pre-print
AnoNet can learn from a limited number of images. For one of the data-sets, AnoNet learnt to detect anomalies after a single pass through just 53 training images.  ...  Today, manual human visual inspection is still the norm because it is difficult to automate anomaly detection. Neural networks are a useful tool that can teach a machine to find defects.  ...  Currently, there is no bench-marking available for weakly supervised anomaly detection. We hope that AnoNet serves as a benchmark for future studies.  ... 
arXiv:1911.10608v1 fatcat:eh7xu2lrizf6zb555obsmaujhi

Deep Convolution Neural Networks for Image Classification

Arun D. Kulkarni
2022 International Journal of Advanced Computer Science and Applications  
DCNN models such as Alex Net, VGG Net, and Google Net have been used to classify large dataset having millions of images into thousand classes.  ...  In the first and second dataset seventy percent randomly chosen samples from each class were used for training.  ...  ACKNOWLEDGMENT The author is thankful to UT Health North, Tyler, TX for providing pleura image dataset.  ... 
doi:10.14569/ijacsa.2022.0130603 fatcat:rktjxiakn5f6rpduocb2dfbnk4

SUPERVISED CLASSIFICATION METHODS FOR AUTOMATIC DAMAGE DETECTION CAUSED BY HEAVY RAINFALL USING MULTITEMPORAL HIGH RESOLUTION OPTICAL IMAGERY AND AUXILIARY DATA

A. Cerbelaud, L. Roupioz, G. Blanchet, P. Breil, X. Briottet
2021 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
frequency optical imagery (Sentinel-2, Pléiades) are used to develop an automatic damage detection method based on supervised classification algorithms.  ...  For decades now, remote sensing has been largely used to investigate spatial and temporal changes in land use and water resources.  ...  A plot-based supervised classification method was thus achieved relying on the combined search for spectral, temporal and spatial variations in Sentinel-2 change images acquired only twenty days apart  ... 
doi:10.5194/isprs-archives-xliii-b3-2021-693-2021 fatcat:nkyesltl6zeydjpneriosqb5xe

LSTD: A Low-Shot Transfer Detector for Object Detection

Hao Chen, Yali Wang, Guoyou Wang, Yu Qiao
2018 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
knowledge respectively from source and target domains, in order to further enhance fine-tuning with a few target images.  ...  Recent advances in object detection are mainly driven by deep learning with large-scale detection benchmarks.  ...  One popular solution is to collect extra detection images but with easily-annotated labels (e.g., image-level supervision).  ... 
doi:10.1609/aaai.v32i1.11716 fatcat:dbyyiurwaje35pf5fjrwytt7pu

EZLearn: Exploiting Organic Supervision in Automated Data Annotation

Maxim Grechkin, Hoifung Poon, Bill Howe
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
supervised methods trained on tens of thousands of annotated examples.  ...  Distant supervision has emerged as a promising paradigm for exploiting such indirect supervision by automatically annotating examples where the text description contains a class mention in the lexicon.  ...  In functional genomics, there are thousands of relevant classes. In scientific figure comprehension, prior work only considers three coarse classes, which we expand to twenty-four.  ... 
doi:10.24963/ijcai.2018/568 dblp:conf/ijcai/GrechkinPH18 fatcat:sq6vmhjy3zf5ph3yvh6qwiuzke

What can Machine Learning do for Radio Spectrum Management?

Ebtesam Almazrouei, Gabriele Gianini, Nawaf Almoosa, Ernesto Damiani
2020 Proceedings of the 16th ACM Symposium on QoS and Security for Wireless and Mobile Networks  
Twenty-ve thousands of data are used for training, and ve thousand are used for testing. Results show that regression performs better than ML classi cation methods.  ...  Nowadays, DL models outperform human levels of accuracy of detection algorithms for objects in images or handwritten digits [47] .  ... 
doi:10.1145/3416013.3426443 dblp:conf/mswim/AlmazroueiGAD20 fatcat:2hjionukbjgo5gx6bnknaiymse

BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning [article]

Fisher Yu, Haofeng Chen, Xin Wang, Wenqi Xian, Yingying Chen, Fangchen Liu, Vashisht Madhavan, Trevor Darrell
2020 arXiv   pre-print
The dataset possesses geographic, environmental, and weather diversity, which is useful for training models that are less likely to be surprised by new conditions.  ...  We construct BDD100K, the largest driving video dataset with 100K videos and 10 tasks to evaluate the exciting progress of image recognition algorithms on autonomous driving.  ...  We use Mask R-CNN [16] with ResNet-50 [17] as the backbone, and train detection and instance segmentation in a batch-level round-robin manner.  ... 
arXiv:1805.04687v2 fatcat:7q4hsr2435eqzlz4rzienja2vm

Early Detection of Diabetic Retinopathy

Vamsi Krishna Mekala
2021 International Journal for Research in Applied Science and Engineering Technology  
The classifier is constructed using CNN.  ...  Diagnostic imaging is an important aspect of medical photography in contemporary world. Deep learning improves the eyesight for identifying illness in radiography.  ...  IMAGE CLASSIFICATION A technique taught to identify a supervised learning method (items to categorise in images) using tagged exemplar photographs is known as visual categorization.  ... 
doi:10.22214/ijraset.2021.37918 fatcat:kj4lks76lvclni6ngbdpiz4ag4

Deep ViT Features as Dense Visual Descriptors [article]

Shir Amir, Yossi Gandelsman, Shai Bagon, Tali Dekel
2021 arXiv   pre-print
We demonstrate that such features, when extracted from a self-supervised ViT model (DINO-ViT), exhibit several striking properties: (i) the features encode powerful high level information at high spatial  ...  These properties allow us to design powerful dense ViT descriptors that facilitate a variety of applications, including co-segmentation, part co-segmentation and correspondences -- all achieved by applying  ...  number of input images, ranging from as little as a pair of images to a collection containing thousands of images.  ... 
arXiv:2112.05814v1 fatcat:f5gndkdx45gvtjoqo7peqh4xvy

Determining the Optimum Maturity of Maize Using Computational Intelligence Techniques

Ayuba Peter, Luhutyit Peter Damuut, Sa'adatu Abdulkadir
2020 American Journal of Neural Networks and Applications  
This paper leverages on the use of Artificial Neural Networks (ANN) interfaced with image processing and Convolutional Neural Networks (pre-trained ResNet50 Network) in determining the optimum ripeness  ...  Therefore, this research posits that farmers could be sensitized on the possibility of utilizing image processing and neural networks technique in the determination of the maturity of maize in the nearest  ...  In the case of the texture feature, Grey Level Co-occurrence Matrix (GLCM) is used to create a gray-level co-occurrence matrix from the images.  ... 
doi:10.11648/j.ajnna.20200601.11 fatcat:af5cnjsx3vbs7i2mihrgswpxcy
« Previous Showing results 1 — 15 out of 16,821 results