Et al

By P. Cochat, L. Vaucoret, J. Sarles. 2008

75,403 citations recorded by the source corpus.

Subfields: Sociology and Political Science, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Sociology and Political Science

In Evaluation & Benchmarks

Citations in the corpus

Deep Residual Learning for Image Recognition (2015, methodology): “ImageNet Classification We evaluate our method on the ImageNet 2012 classification dataset [35] that consists of 1000 classes.”
Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification (2015, methodology): “1% in [26]) of a dedicated individual labeler on this recognition challenge.”
Convolutional neural networks at constrained time cost (2014, methodology): “We compare our models with the existing models on the 1000-category ImageNet 2012 dataset [3, 20].”
Direct Preference Optimization: Your Language Model is Secretly a Reward Model (2023, background): “Large unsupervised language models (LMs) trained on very large datasets acquire surprising capabilities [11, 7, 37, 8].”
Adaptive Computation Time for Recurrent Neural Networks (2016, neutral): “The evidence that increased depth leads to more performant networks is by now inarguable [5, 4, 19, 9], and recent results show that increased sequence length can be similarly beneficial [31, 33, 25].”
Snorkel: Rapid Training Data Creation with Weak Supervision (2017, methodology): “Both models are implemented in Tensorflow [2] and trained using the Adam optimizer [24], with hyperparameters selected via random grid search using a small labeled development set.”
TruthfulQA: Measuring How Models Mimic Human Falsehoods (2021, methodology): “Another example is when GPT-3 models trained on GitHub learn to produce buggy code [6].”
Scaling Laws for Transfer (2021, background): “We believe it is particularly important to characterize fine-tuning in the low data regime because many tasks of interest won’t have a sufficiently big, readily available dataset to train large models from-scratch (either billions of data points [HNA17, KMH20, HKK20] or a perfect simulator like in G”
A Retrieve-and-Edit Framework for Predicting Structured Outputs (2018, methodology): “In our experiments, we let pedit be a standard sequence-to-sequence model with attention and copying [12, 36] (see Appendix B for details), but any model architecture can be used for the editor.”
Efficient Natural Language Response Suggestion for Smart Reply (2017, methodology): “The models are trained on CPUs across 50 machines using a distributed implementation of TensorFlow [1].”
Data-Efficient Image Recognition with Contrastive Predictive Coding (2019, methodology): “As in [49], we trained a CPC feature extractor on the ILSVRC ImageNet competition dataset images [58].”
Local Relation Networks for Image Recognition (2019, methodology): “Appearance composability We follow a general approach for relation modeling [1] to compute appearance composability Φ(fθq (xp′), fθk(xp)), where xp′ and xp are projected to a query (by a channel transformation layer fθq ) and key (by a channel transformation layer fθk ) embedding space, respectively”
Attention Augmented Convolutional Networks (2019, background): “For example, improving the “backbone” network typically leads to improvements in object detection [19] and image segmentation [6].”
Weight Standardization (2019, neutral): “geNet rather than Kinetics dataset [18] as the starting points.”
Group Normalization (2018, methodology): “its BN counterpart on Mask R-CNN for COCO object detection and segmentation [37], and on 3D convolutional networks for Kinetics video classification [30].”
Self-Supervised MultiModal Versatile Networks (2020, methodology): “Taking this idea into the self-supervised domain [8], multiple works use a pretext task of predicting whether visual and audio signals come from the same video [8, 9, 44, 58, 61, 72].”
A Short Note on the Kinetics-700 Human Action Dataset (2019, background): “The additional classes for Kinetics-700 over Kinetics600 were partly sourced from the lists of actions (or verbs) in recent human action datasets, such as EPIC-Kitchens [4] and AVA [5].”
HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips (2019, background): “Charades [42] 10k 16k 10,000 82h Home 2016 MSR-VTT [52] 10k 200k 7,180 40h Youtube 2016 YouCook2 [61] 14k 14k 2,000 176h Youtube 2018 EPIC-KITCHENS [5] 40k 40k 432 55h Home 2018 DiDeMo [11] 27k 41k 10,464 87h Flickr 2017 M-VAD [46] 49k 56k 92 84h Movies 2015 MPII-MD [37] 69k 68k 94 41h Movies 2015 A”
Towards VQA Models That Can Read (2019, methodology): “We use Open Images v3 dataset [27] as the source of our images.”