Pascal consists oftwoparts: astandard train16 set ('train') and a train set labeled with a different distribution of human annotations ('trainaug').

We provide a detailed description of all tasks and some additional details on the training of MDEQ. The entire dataset is divided into training (50K images) and testing (10K) sets. We use two different training settings for evaluating the MDEQ model on CIFAR-10. In the second setting, we apply data augmentation to the input images (i.e., The dataset we use contains 1.2 million labeled training images from ImageNet [ Each pixel is classified in a 19-way fashion for evaluation. CIFAR-10 classification models were trained on 1 GPU (including the baselines).

Neural Information Processing Systems http://nips.cc/

Both RTFormer-Slim and RTFormer-Base outperform the corresponding DDRNet variants. The self-attention used for comparison is following (12). For linformer attention, we directly give a result without hyper parameter modification. Multi-head external attention can achieve a good inference speed, which is benefit from its linear complexity and the design of sharing external parameter for multiple heads. "#Params" refers to the number of parameters.

The paper proposes channel-wise convolutions that address the full connections between feature maps and replace them with sparse connections (based on 1-D convolutions). This reduces the #params and #FLOPS significantly; while maintaining high accuracy. The authors show results on imagenet classification and compare it to VGG/MobileNet variants to demonstrate this. Strengths: The paper is well written and easy to follow. Background and related work such as standard convolution fc layers used in neural nets; mobilenet and shufflenet variants to reduce computation are described in sufficient detail.

Generative training has been demonstrated to be powerful for building visual-language models. However, on zero-shot discriminative benchmarks, there is still a performance gap between models trained with generative and discriminative objectives. In this paper, we aim to narrow this gap by improving the efficacy of generative training on classification tasks, without any finetuning processes or additional modules. Specifically, we focus on narrowing the gap between the generative captioner and the CLIP classifier. We begin by analysing the predictions made by the captioner and classifier and observe that the caption generation inherits the distribution bias from the language model trained with pure text modality, making it less grounded on the visual signal. To tackle this problem, we redesign the scoring objective for the captioner to alleviate the distributional bias and focus on measuring the gain of information brought by the visual inputs. We further design a generative training objective to match the evaluation objective. We name our model trained and evaluated from the novel procedures as Information Gain (IG) captioner. We pretrain the models on the public Laion-5B dataset and perform a series of discriminative evaluations. For the zero-shot classification on ImageNet, IG captioner achieves $> 18\%$ improvements over the standard captioner, achieving comparable performances with the CLIP classifier. IG captioner also demonstrated strong performance on zero-shot image-text retrieval tasks on MSCOCO and Flickr30K. We hope this paper inspires further research towards unifying generative and discriminative training procedures for visual-language models.

We trained a large, deep convolutional neural network to classify the 1.3 million high-resolution images in the LSVRC-2010 ImageNet training set into the 1000 different classes. On the test data, we achieved top-1 and top-5 error rates of 39.7\% and 18.9\% which is considerably better than the previous state-of-the-art results. The neural network, which has 60 million parameters and 500,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and two globally connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implementation of convolutional nets. To reduce overfitting in the globally connected layers we employed a new regularization method that proved to be very effective.

Recent progress in image recognition has stimulated the deployment of vision systems at an unprecedented scale. As a result, visual data are now often consumed not only by humans but also by machines. Existing image processing methods only optimize for better human perception, yet the resulting images may not be accurately recognized by machines. This can be undesirable, e.g., the images can be improperly handled by search engines or recommendation systems. In this work, we examine simple approaches to improve machine recognition of processed images: optimizing the recognition loss directly on the image processing network or through an intermediate input transformation model. Interestingly, the processing model's ability to enhance recognition quality can transfer when evaluated on models of different architectures, recognized categories, tasks and training datasets. This makes the methods applicable even when we do not have the knowledge of future recognition models, e.g., when uploading processed images to the Internet. We conduct experiments on multiple image processing tasks paired with ImageNet classification and PASCAL VOC detection as recognition tasks. With these simple yet effective methods, substantial accuracy gain can be achieved with strong transferability and minimal image quality loss. Through a user study we further show that the accuracy gain can transfer to a black-box cloud model. Finally, we try to explain this transferability phenomenon by demonstrating the similarities of different models' decision boundaries. Code is available at https://github.com/liuzhuang13/Transferable_RA .