WildWood: a new Random Forest algorithm

Github: https://github.com/pyensemble/wildwood

Paper: https://arxiv.org/abs/2109.08010v1

Documentation: https://wildwood.readthedocs.io/

👉 @bigdata_1

Textual Entailment for Event Argument Extraction: Zero- and Few-Shot with Multi-Source Learning

Github: https://github.com/osainz59/Ask2Transformers

Instructions: https://github.com/facebookresearch/metaseq/blob/main/docs/setup.md

Paper: https://arxiv.org/abs/2205.01376v1

Dataset: https://paperswithcode.com/dataset/snli

👉 @bigdata_1

Simple Entity-Centric Questions Challenge Dense Retrievers

Github: https://github.com/princeton-nlp/entityquestions

Paper: https://arxiv.org/abs/2109.08535v1

Dataset: https://paperswithcode.com/dataset/natural-questions

👉 @bigdata_1

Textless NLP: Generating expressive speech from raw audio

Facebook Ai: https://ai.facebook.com/blog/textless-nlp-generating-expressive-speech-from-raw-audio/

Examples: https://speechbot.github.io/pgslm/

Code: https://github.com/pytorch/fairseq/tree/master/examples/textless_nlp/gslm

Paper: https://arxiv.org/abs/2102.01192

👉 @bigdata_1

Unified Data Collection for Visual-Inertial Calibration via Deep Reinforcement Learning

Github: https://github.com/ethz-asl/Learn-to-Calibrate

Paper: https://arxiv.org/abs/2109.14974v1

👉 @bigdata_1

Dynamic Slimmable Network (DS-Net)

Github: https://github.com/M3DV/AlignShift

Paper: https://arxiv.org/abs/2109.10060v1

👉 @bigdata_1

EasyNLP is an easy-to-use NLP development and application toolkit in PyTorch

$ pip install pai-easynlp.

Github: https://github.com/alibaba/EasyNLP

Paper: https://arxiv.org/abs/2205.03071v1

Dataset: https://paperswithcode.com/dataset/natural-questions

Documentation: https://www.yuque.com/easyx/easynlp/kkhkai

👉 @bigdata_1

Awakening Latent Grounding from Pretrained Language Models for Semantic Parsing

Github: https://github.com/microsoft/ContextualSP

Paper: https://arxiv.org/abs/2109.10540v1

👉 @bigdata_1

DeepFilterNet

A Low Complexity Speech Enhancement Framework for Full-Band Audio (48kHz) using on Deep Filtering.

Github: https://github.com/rikorose/deepfilternet

Paper: https://arxiv.org/abs/2205.05474v1

Demo: https://huggingface.co/spaces/hshr/DeepFilterNet2

👉 @bigdata_1

Hierarchical Memory Matching Network for Video Object Segmentation

Github: https://github.com/hongje/hmmn

Paper: https://arxiv.org/abs/2109.11404v1

Dataset: https://paperswithcode.com/dataset/davis-2016

👉 @bigdata_1

KG-SP: Knowledge Guided Simple Primitives for Open World Compositional Zero-Shot Learning

Github: https://github.com/explainableml/kg-sp

Paper: https://arxiv.org/abs/2205.06784v1

Dataset: https://paperswithcode.com/dataset/conceptnet

👉 @bigdata_1

Deep Social Force

Github: https://github.com/svenkreiss/socialforce

Paper: https://arxiv.org/abs/2109.12081v1

👉 @bigdata_1

OWL-ViT: Open-World Object Detection with Vision Transformers

OWL-ViT is an open-vocabulary object detector.

Github: https://github.com/google-research/scenic/tree/main/scenic/projects/owl_vit

Paper: https://arxiv.org/abs/2205.06230

Colab: https://colab.research.google.com/github/google-research/scenic/blob/main/scenic/projects/owl_vit/notebooks/OWL_ViT_minimal_example.ipynb

Dataset: https://paperswithcode.com/dataset/objects365

👉 @bigdata_1

TSM: Temporal Shift Module for Efficient Video Understanding

Github: https://github.com/MIT-HAN-LAB/temporal-shift-module

Paper: https://arxiv.org/abs/1811.08383

Demo: https://www.youtube.com/watch?v=0T6u7S_gq-4

👉 @bigdata_1

AdaVAE: Exploring Adaptive GPT-2s in Variational Auto-Encoders for Language Modeling

First VAE framework empowered with adaptive GPT-2s (AdaVAE).

Github: https://github.com/ImKeTT/adavae

Paper: https://arxiv.org/abs/2205.05862v1

Task: https://paperswithcode.com/task/representation-learning

👉 @bigdata_1

TSM: Temporal Shift Module for Efficient Video Understanding

Github: https://github.com/nikitakit/self-attentive-parser

Paper: https://arxiv.org/abs/2109.12814v1

Dataset: https://paperswithcode.com/dataset/penn-treebank

👉 @bigdata_1

Efficient and performance-portable vector software

CPUs provide SIMD/vector instructions that apply the same operation to multiple data items. This can reduce energy usage e.g. fivefold because fewer instructions are executed. We also often see 5-10x speedups.

Code: https://github.com/google/highway

Paper: https://arxiv.org/abs/2205.05982v1

Testing: https://github.com/google/highway/blob/master/g3doc/release_testing_process.md

👉 @bigdata_1

Towards Flexible Blind JPEG Artifacts Removal

Github: https://github.com/jiaxi-jiang/fbcnn

Paper: https://arxiv.org/abs/2109.14573v1

👉 @bigdata_1

[RK-Net]Joint Representation Learning and Keypoint Detection for Cross-view Geo-localization

Github: https://github.com/AggMan96/RK-Net

Paper: https://zhunzhong.site/paper/RK_Net.pdf

Dataset: https://paperswithcode.com/dataset/university-1652

👉 @bigdata_1

Evaluation of Deep Neural Network Domain Adaptation Techniques for Image Recognition

Github: https://github.com/agrija9/deep-unsupervised-domain-adaptation

Paper: https://arxiv.org/abs/2109.13420v1

Dataset: https://paperswithcode.com/dataset/office-31

👉 @bigdata_1

Group R-CNN for Point-based Weakly Semi-supervised Object Detection

Instance-aware representation learning which consists of instance-aware feature enhancement and instance-aware parameter generation to overcome this issue.

Code: https://github.com/jshilong/grouprcnn

Installation: https://mmdetection.readthedocs.io/en/v2.18.1/get_started.html#installation

Paper: https://arxiv.org/abs/2205.05920v1

Task: https://paperswithcode.com/task/object-detection

Dataset: https://paperswithcode.com/dataset/coco

👉 @bigdata_1