Agent that follows a pre-defined action sequence

Visualizer that draws data about detection/segmentation on images. It contains methods like `draw_{text,box,circle,line,binary_mask,polygon}

Represents a unique interactable object in the house.

Check whether the contents appear in the page

Stores meta information for an object

The goal of this environment is to produce a prototype of a browser environment. In the end, we want to support a fully configurable browser

Check whether the answer is correct by querying a vision model.

Fuses the information from the Fast pathway to the Slow pathway. Given the tensors from Slow pathway and Fast pathway, fuse information from

Fuses the information from the Fast pathway to the Slow pathway. Given the tensors from Slow pathway and Fast pathway, fuse information from

Attribute: polygons (list[ndarray]): list[ndarray]: polygons for this mask. Each ndarray has format [x, y, x, y, ...]

Attribute: polygons (list[ndarray]): list[ndarray]: polygons for this mask. Each ndarray has format [x, y, x, y, ...]

A cache for image features. These features might be reused at different parts of the inference pipeline. This class provide a easy interf

Check whether the URL is exactly the same as of the reference URLs

a few conv layers to flatten features that come out of ResNet

LabeledVideoDataset handles the storage, loading, decoding and clip sampling for a video dataset. It assumes each video is stored as either a

LabeledVideoDataset handles the storage, loading, decoding and clip sampling for a video dataset. It assumes each video is stored as either a

Track a series of values and provide access to smoothed values over a window or the global series average.

Check whether the answer is correct with: exact match: the answer is exactly the same as the reference answer must include: each phrase in the

Works for DAVIS/YouTubeVOS/BL30K training For each sequence: - Pick num_frames frames - Pick max_num_obj objects - Apply some ran

The goal of this environment is to produce a prototype of a browser environment. In the end, we want to support a fully configurable browser

Samples just a single clip from the center of the video (use for testing)

Sampler that restricts data loading to a subset of the dataset. It is especially useful in conjunction with :class:`torch.nn.parallel.Distribu

Works for key/value pairs type storage e.g., working and long-term memory

LLM Planner for going from teach dialogue to executable program

This is a 2D attention module, which only returns the attention softmax (no multiplication by value)

Very simple multi-layer perceptron (also called FFN)

Very simple multi-layer perceptron (also called FFN)

Simple convolutional head, using group norm. Upsampling is done using a FPN approach

Taken from DINO # Copyright (c) Facebook, Inc. and its affiliates. # # Licensed under the Apache License, Version 2.0 (the "License"

Transformer-style positional encoding with wavelets

This class converts the output of the model to the final panoptic result, in the format expected by the coco panoptic API