Fahim Farook
Posts
1639Following
139Followers
885I'm currently working on my second novel which is complete, but is in the edit stage. I wrote my first novel over 20 years ago but then didn't write much till now.
I post about #Coding, #Flutter, #Writing, #Movies and #TV. I'll also talk about #Technology, #Gadgets, #MachineLearning, #DeepLearning and a few other things as the fancy strikes ...
Lived in: ๐ฑ๐ฐ๐ธ๐ฆ๐บ๐ธ๐ณ๐ฟ๐ธ๐ฌ๐ฒ๐พ๐ฆ๐ช๐ซ๐ท๐ช๐ธ๐ต๐น๐ถ๐ฆ๐จ๐ฆ
Fahim Farook
f
"The Infinite Index: Information Retrieval on Generative Text-To-Image Models. (arXiv:2212.07476v2 [cs.IR] UPDATED)" โ A reframing of prompt engineering for generative models as interactive text-based retrieval on a novel kind of "infinite index".
Paper: http://arxiv.org/abs/2212.07476
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Overview of indexing approachesโฆ
Paper: http://arxiv.org/abs/2212.07476
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Overview of indexing approachesโฆ
0
3
0
Fahim Farook
f
"Look at Adjacent Frames: Video Anomaly Detection without Offline Training. (arXiv:2207.13798v5 [cs.CV] UPDATED)" โ A solution to detect anomalous events in videos without the need to train a model offline, based on a randomly-initialized multilayer perceptron that is optimized online to reconstruct video frames, pixel-by-pixel, from their frequency information.
Paper: http://arxiv.org/abs/2207.13798
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Two figures, the first show theโฆ
Paper: http://arxiv.org/abs/2207.13798
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Two figures, the first show theโฆ
0
1
0
Fahim Farook
f
"Synthesis of Compositional Animations from Textual Descriptions. (arXiv:2103.14675v6 [cs.CV] UPDATED)" โ A technique for generating compositional actions, which handles complex input sentences and outputs a 3D pose sequence depicting the actions in the input sentence.
Paper: http://arxiv.org/abs/2103.14675
Code: https://github.com/anindita127/complextext2animation
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Overview of our proposed methodโฆ
Paper: http://arxiv.org/abs/2103.14675
Code: https://github.com/anindita127/complextext2animation
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Overview of our proposed methodโฆ
0
1
0
Fahim Farook
f
"Optimising Event-Driven Spiking Neural Network with Regularisation and Cutoff. (arXiv:2301.09522v1 [cs.CV])" โ A proposal to strengthen the marriage between SNNs and event-based inputs by considering anytime optimal inference SNNs, or AOI-SNNs, which can terminate anytime during the inference to achieve optimal inference result.
Paper: http://arxiv.org/abs/2301.09522
Code: https://github.com/Dengyu-Wu/SNN-Regularisation-Cutoff
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
An illustrative diagram showingโฆ
Paper: http://arxiv.org/abs/2301.09522
Code: https://github.com/Dengyu-Wu/SNN-Regularisation-Cutoff
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
An illustrative diagram showingโฆ
0
1
0
Fahim Farook
f
"OvarNet: Towards Open-vocabulary Object Attribute Recognition. (arXiv:2301.09506v1 [cs.CV])" โ An examination of the problem of simultaneously detecting objects and inferring their visual attributes in an image, even for those with no manual annotations provided at the training stage, resembling an open-vocabulary scenario.
Paper: http://arxiv.org/abs/2301.09506
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
The first row depicts the tasksโฆ
Paper: http://arxiv.org/abs/2301.09506
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
The first row depicts the tasksโฆ
0
1
1
Fahim Farook
f
"RainDiffusion:When Unsupervised Learning Meets Diffusion Models for Real-world Image Deraining. (arXiv:2301.09430v1 [cs.CV])" โ An unsupervised image deraining paradigm based on diffusion models which introduces stable training of unpaired real-world data instead of weak adversarial training.
Paper: http://arxiv.org/abs/2301.09430
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Image deraining results on a reโฆ
Paper: http://arxiv.org/abs/2301.09430
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Image deraining results on a reโฆ
0
1
1
Fahim Farook
f
"Computer Vision for a Camel-Vehicle Collision Mitigation System. (arXiv:2301.09339v1 [cs.CV])" โ Testing different object detection models on the task of detecting camels on the road since in Saudi Arabia, due to the size of camels, camel-vehicle collisions result in a 25% fatality rate.
Paper: http://arxiv.org/abs/2301.09339
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Example images from the datasetโฆ
Paper: http://arxiv.org/abs/2301.09339
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Example images from the datasetโฆ
0
1
0
Fahim Farook
f
"Efficient Training Under Limited Resources. (arXiv:2301.09264v1 [cs.LG])" โ A demonstration that Neural Architecture Search (NAS), Hyper Parameters Optimization (HPO), and Data Augmentation help Deep Neural Networks (DNNs) perform much better when training time budget and size of the dataset are limited.
Paper: http://arxiv.org/abs/2301.09264
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Paper: http://arxiv.org/abs/2301.09264
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
0
2
1
Fahim Farook
f
"Apples and Oranges? Assessing Image Quality over Content Recognition. (arXiv:2301.09190v1 [cs.CV])" โ An investigation of whether image recognition and quality assessment can be performed in a multitask learning manner.
Paper: http://arxiv.org/abs/2301.09190
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Architecture of the proposed IQโฆ
Paper: http://arxiv.org/abs/2301.09190
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Architecture of the proposed IQโฆ
0
2
0
Fahim Farook
f
"Raw or Cooked? Object Detection on RAW Images. (arXiv:2301.08965v1 [cs.CV])" โ An investigation of the hypothesis that the intermediate representation of visually pleasing images is sub-optimal for downstream computer vision tasks compared to the RAW image representation.
Paper: http://arxiv.org/abs/2301.08965
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Three qualitative examples fromโฆ
Paper: http://arxiv.org/abs/2301.08965
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Three qualitative examples fromโฆ
0
3
2
@greysemanticist I used to use the Mycroft extension to add custom search engines to Firefox, but have basically not used custom engines for at least 3 - 4 years and so have no idea if Mycroft still works and/or if it even exists anymore โฆ
1
0
0
Fahim Farook
f
"Enactive Artificial Intelligence: Overcoming Male Gaze in Robot-Human Interaction" โ The paper characterises AI design as a cultural practice; which is then specified in feminist technoscience principles, i.e. how gender and other embodied identity markers are entangled in AI. From this, the paper derives the conditions for Enactive Artificial Intelligence.
Paper: https://arxiv.org/abs/2301.08741
#AI #NewPaper #Robotics #HumanComputerInteraction
<<Find this useful? Please boost so that others can benefit too ๐>>
Paper: https://arxiv.org/abs/2301.08741
#AI #NewPaper #Robotics #HumanComputerInteraction
<<Find this useful? Please boost so that others can benefit too ๐>>
0
1
0
Fahim Farook
f
"Is ChatGPT A Good Translator? A Preliminary Study" โ An evaluation of ChatGPT for machine translation, including translation prompt, multilingual translation, and translation robustness.
TL;DR; "ChatGPT performs competitively with commercial translation products (e.g., Google Translate) on high-resource European languages but lags behind significantly on low-resource or distant languages."
Paper: https://arxiv.org/abs/2301.08745
#AI #NewPaper #DeepLearning #MachineLearning #Language
<<Find this useful? Please boost so that others can benefit too ๐>>
TL;DR; "ChatGPT performs competitively with commercial translation products (e.g., Google Translate) on high-resource European languages but lags behind significantly on low-resource or distant languages."
Paper: https://arxiv.org/abs/2301.08745
#AI #NewPaper #DeepLearning #MachineLearning #Language
<<Find this useful? Please boost so that others can benefit too ๐>>
0
1
0
Fahim Farook
f
"Domain-agnostic and Multi-level Evaluation of Generative Models" โ A framework for multi-level performance evaluation of generative models which could be employed across different domains (images, text, graphs, molecules, etc.)
Paper: https://arxiv.org/abs/2301.08750
#AI #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Overview of the MPEGO frameworkโฆ
Paper: https://arxiv.org/abs/2301.08750
#AI #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Overview of the MPEGO frameworkโฆ
0
1
0
Fahim Farook
f
@Herstory We did try story books in the early days (3 - 4 years ago) but at that point we were trying to run before we could walk ๐
Now would certainly be a good time to go back to them, but personally, I'm too impatient to do that if I'm busy with other stuff. I hate having to stop to look up words instead of gaining an idea of what I'm reading based on context. My wife is more patient ๐
So we probably will try the comics for a while and then go on from there. Thank you for the suggestions!
Now would certainly be a good time to go back to them, but personally, I'm too impatient to do that if I'm busy with other stuff. I hate having to stop to look up words instead of gaining an idea of what I'm reading based on context. My wife is more patient ๐
So we probably will try the comics for a while and then go on from there. Thank you for the suggestions!
0
0
0
Fahim Farook
f
"Improving Deep Regression with Ordinal Entropy. (arXiv:2301.08915v1 [cs.CV])" โ An investigation of the fact that in computer vision, formulating regression problems as a classification task often yields better performance, and shows that classification, with the cross-entropy loss, outperforms regression with a mean squared error loss in its ability to learn high-entropy feature representations.
Paper: http://arxiv.org/abs/2301.08915
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Feature learning of regression โฆ
Paper: http://arxiv.org/abs/2301.08915
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Feature learning of regression โฆ
0
1
2
Fahim Farook
f
"A Large-scale Film Style Dataset for Learning Multi-frequency Driven Film Enhancement. (arXiv:2301.08880v1 [cs.CV])" โ A large-scale and high-quality film style dataset to facilitate film-based image stylization research. The dataset includes three different film types and more than 5000 in-the-wild high resolution images.
Paper: http://arxiv.org/abs/2301.08880
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
This figure contains input imagโฆ
Paper: http://arxiv.org/abs/2301.08880
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
This figure contains input imagโฆ
0
1
0
Fahim Farook
f
"Regeneration Learning: A Learning Paradigm for Data Generation. (arXiv:2301.08846v1 [cs.LG])" โ A learning paradigm for data generation (e.g., text generation, speech recognition, speech synthesis, music composition, image generation, and video generation) which first generates Y' (an abstraction/representation of the target data, Y) from the source data, X, and then generates Y from Y'.
Paper: http://arxiv.org/abs/2301.08846
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Three types of tasks in machineโฆ
Paper: http://arxiv.org/abs/2301.08846
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Three types of tasks in machineโฆ
0
1
0
Fahim Farook
f
"In-situ Water quality monitoring in Oil and Gas operations. (arXiv:2301.08800v1 [cs.CV])" โ a model designed to enable users to determine contamination levels in water bodies with weak reflectance patterns such as small ponds based on satellite images.
Paper: http://arxiv.org/abs/2301.08800
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Image โaโshows data sample fromโฆ
Paper: http://arxiv.org/abs/2301.08800
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Image โaโshows data sample fromโฆ
0
1
0
Fahim Farook
f
"Visual Semantic Relatedness Dataset for Image Captioning. (arXiv:2301.08784v1 [cs.CL])" โ A textual visual context dataset for captioning, in which the publicly available dataset COCO Captions has been extended with information about the scene (such as objects in the image).
Paper: http://arxiv.org/abs/2301.08784
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Examples of our proposed COCO bโฆ
Paper: http://arxiv.org/abs/2301.08784
#AI #CV #NewPaper #DeepLearning #MachineLearning
<<Find this useful? Please boost so that others can benefit too ๐>>
Examples of our proposed COCO bโฆ
0
1
0