Rowel Atienza 🇵🇭

Creator of ViTSTR and EfficientSpeech (ICASSP2023) and co-creator of PARSeq. Professor & Scientist at the University of the Philippines.

[Rowel Atienza 🇵🇭 @jacobe]

AI agents shouldn't be complex, heavy, or costly. baby-onit: one file, ~1,000 lines. Everything is traceable. Use @tavilyai and @OpenRouter or @vercel to start using agents without incurring initial costs. https://t.co/MIo66MsdX5

[Rowel Atienza 🇵🇭 @jacobe]

@Em_alej HiFiGAN

[Rowel Atienza 🇵🇭 @jacobe]

Most ARM chips can't run decent AI models. Introducing EfficientSpeech, a 266k-param TTS model. Low cost ARM chips like in RPi4 can generate 104sec of speech mel spec in 1sec. Here's an AI-generated video w/ voice from EfficientSpeech. Info: https://t.co/z0ZzZHZEMH #ICASSP2023

[Rowel Atienza 🇵🇭 @jacobe]

We are lucky to have this paper accepted at ECCV 2022! More details later ... #ECCV2022

[DeepAI @DeepAI]

Scene Text Recognition with Permuted Autoregressive Sequence Models
https://t.co/9VG7mdgG4a
by Darwin Bautista and @jacobe
#Computation #Language

[Rowel Atienza 🇵🇭 @jacobe]

@arXiv_Daily Simple yet effective idea: Remove inefficient top-most layers & replace them with an efficient head. For VWW, param count reduced by 93% with only 0.65% accuracy decrease. Counterintuitively, the quantized pruned net increased its accuracy on ARM Cortex M0.

[arXiv Daily @Arxiv_Daily]

Depth Pruning with Auxiliary Networks for TinyML
https://t.co/40QhtcujFG
by Josen Daniel De Leon and @jacobe
#NeuralNetwork #ComputerScience

[AK @_akhaliq]

.@Gradio & @huggingface for Rapid Deep Learning App Development by @jacobe
link: https://t.co/R6IqNQSZTw

[Rowel Atienza 🇵🇭 @jacobe]

@ak92501 @Gradio @huggingface Thanks. Let's keep on building better infrastructure and tooling to make AI more accessible.

[University of the Philippines @upsystem]

"The UP National Engineering Center Analytics and Data Science Certifications announced the development on Wednesday." https://t.co/UQ7QXEe2v9 via @gmanews

[Rowel Atienza 🇵🇭 @jacobe]

Idea: If data augmentation improves model generalization, why not use it to generate 2 new inputs and force the representations to agree. Result: Additional model performance improvement. Comparison: Unlike Label Smoothing, the performance of our method, AgMax, is consistent.

[DeepAI @DeepAI]

Improving Model Generalization by Agreement of Learned Representations from Data Augmentation
https://t.co/QuQDB9MWzW
by @jacobe
#ComputerVision #ImageNet

[Tim Dettmers @Tim_Dettmers]

I am excited to share my latest work: 8-bit optimizers – a replacement for regular optimizers. Faster 🚀, 75% less memory 🪶, same performance📈, no hyperparam tuning needed 🔢. 🧵/n

Paper: https://t.co/V5tjOmaWvD
Library: https://t.co/JAvUk9hrmM
Video: https://t.co/TWCNpCtCap

[AI at Meta @AIatMeta]

We’re introducing GSLM, the first language model that breaks free completely of the dependence on text for training. This “textless NLP” approach learns to generate expressive speech using only raw audio recordings as input. Learn more and get the code:
https://t.co/kRkUaFyZWb

[Rowel Atienza 🇵🇭 @jacobe]

@arXiv_Daily Data Augmentation for STR will be presented at ICCV 2021 Workshop on Interactive Labeling and Data Augmentation for Vision. GitHub: https://t.co/OLfGeLQRfc

[arXiv Daily @Arxiv_Daily]

Data Augmentation for Scene Text Recognition
https://t.co/9J7usHcMut
by @jacobe
#ComputerVision #ComputerVision

[ODSC (Open Data Science Conference) AI @_odsc]

ODSC APAC Virtual Conference Speaker: https://t.co/udawOBNnhi #ODSCAPAC #DataScience #DeepLearning @jacobe @upsystem

[Rowel Atienza 🇵🇭 @jacobe]

My 1st github project that reached the 1k milestone. @PacktPub https://t.co/5zyns8TCbL

[Rowel Atienza 🇵🇭 @jacobe]

@arXiv_Daily It took us more than a year building, collecting, annotating, validating and benchmarking this dataset.
Dataset: https://t.co/t13w9wWlBo
To appear at #CVPR2021 Workshop: https://t.co/H84qAjdSjE

[arXiv Daily @Arxiv_Daily]

GOO: A Dataset for Gaze Object Prediction in Retail Environments
https://t.co/9xGZb2PdfX
by Henri Tomas et al. including @jacobe
#Estimator #Statistics

[Rowel Atienza 🇵🇭 @jacobe]

@Deep__AI Vision Transformer (ViT) for reading real-world text images. My paper will be presented at #ICDAR2021 https://t.co/soUcLttRoy.

[DeepAI @DeepAI]

Vision Transformer for Fast and Efficient Scene Text Recognition
https://t.co/V9xy1WYNcW
by @jacobe
#ComputerVision #PatternRecognition