One-shot viewpoint invariance in matching novel objects

@article{Biederman1999OneshotVI,
  title={One-shot viewpoint invariance in matching novel objects},
  author={Irving Biederman and Moshe Bar},
  journal={Vision Research},
  year={1999},
  volume={39},
  pages={2885-2899}
}

Figures and Tables from this paper

Recognizing novel three–dimensional objects by summing signals from parts and views

  • D. FosterS. Gilson
  • Computer Science
    Proceedings of the Royal Society of London. Series B: Biological Sciences
  • 2002
TLDR
Overall, visual performance proved remarkably lawful: for both long (2 s) and short (100 ms) display durations, it could be summarized by a simple, compact equation with one term representing generalized viewpoint–invariant parts–based processing of 3D object structure, including metric structure, and another term representing structure–invarial processing of 2D views.

Unsupervised invariance learning of transformation sequences in a model of object recognition yields selectivity for non-accidental properties

TLDR
The hypothesis that unsupervised learning of invariances to object transformations may increase the sensitivity to differences in NAPs vs. MPs is tested and it is suggested that greater NAP sensitivity may arise from experiencing different in-depth rotations of objects.

Modulation of Viewpoint Effects in Object Recognition by Shape and Motion Cues

TLDR
It is concluded that coherent motion facilitates generalisation across different views of dynamic objects under some conditions.

Combining disparate views of objects: Viewpoint costs are reduced by stereopsis

An issue of central concern in the object recognition literature is whether changes in the viewpoint from which an object is depicted produces systematic costs in performance, or whether performance

Effect of Active Exploration of 3-D Object Views on the View-Matching Process in Object Recognition

TLDR
The present findings suggest that: (i) active exploration facilitates the process of view matching, (ii) this effect depends on the exploration range and rotational axis, and (iii) interaction between motor and visual information is necessary to facilitate the processof view matching.

Repetition blindness for rotated objects.

TLDR
These findings suggest that RB is due mainly to the activation of object representations from local diagnostic features, but can be modulated by priming on the basis of view similarity.

Matching multicomponent objects from different viewpoints: mental rotation as normalization?

  • B. WillemsJ. Wagemans
  • Psychology
    Journal of experimental psychology. Human perception and performance
  • 2001
TLDR
The effects of viewpoint on performance imply that the matching of objects is viewpoint dependent and requires a process of normalization to undo the differences between the perceptual description and the stored object descriptions, but the lack of a systematic effect of the axis of rotation suggests that this normalization is best understood as not involving a 3-D transformation of stored3-D object models.

Skeletal descriptions of shape provide unique perceptual information for object recognition

TLDR
This work tested whether the human visual system incorporates a three-dimensional skeletal descriptor of shape to determine an object’s identity, and showed that a model of skeletal similarity explained the greatest amount of variance in participants’ object dissimilarity judgments when compared with other computational models of visual similarity.

Object recognition using metric shape

The invariance hypothesis implies domain-specific regions in visual cortex

TLDR
This work can define an index of transformation-compatibility, computable from videos, that can be combined with information about the statistics of natural vision to yield predictions for which object categories ought to have domain-specific regions.
...

References

SHOWING 1-10 OF 43 REFERENCES

Recognizing depth-rotated objects: evidence and conditions for three-dimensional viewpoint invariance.

TLDR
Five experiments on the effects of changes of depth orientation on priming the naming of briefly flashed familiar objects, matching individual sample volumes (geons), and classifying unfamiliar objects all revealed immediate (i.e., not requiring practice) depth invariance.

Testing conditions for viewpoint invariance in object recognition.

  • W. HaywardM. Tarr
  • Psychology
    Journal of experimental psychology. Human perception and performance
  • 1997
TLDR
Results suggest that the conditions proposed by I.C. Biederman and P. Gerhardstein are not generally applicable, the recognition of qualitatively distinct objects often relies on viewpoint-dependent mechanisms, and the molar features of view-based mechanisms appear to be image features rather than geons.

Metric invariance in object recognition: a review and further evidence.

TLDR
Recent priming experiments in which the effects of a prior brief presentation of an image on its subsequent recognition are assessed indicate that the invariance is complete: the magnitude of visual priming is not affected by a change in position, size, orientation in depth, or the particular lines and vertices present in the image, as long as representations of the same components can be activated.

To What Extent Do Unique Parts Influence Recognition Across Changes in Viewpoint?

We investigated how varying the number of unique parts within an object influences recognition across changes in viewpoint The stimuli were shaded objects composed of five three-dimensional volumes

Recognizing novel views of three-dimensional objects.

TLDR
The results indicated that the representation of the objects seen during training was quite viewpoint- specific as recognition of objects in novel orientations was relatively poor, consistent with a growing body of recent research showing that, at least under certain conditions, the visual system stores viewpoint-specific representations of objects.

Recognition-by-components: a theory of human image understanding.

TLDR
Recognition-by-components (RBC) provides a principled account of the heretofore undecided relation between the classic principles of perceptual organization and pattern recognition.

A case of viewer-centered object perception

A network that learns to recognize three-dimensional objects

TLDR
A scheme is developed, based on the theory of approximation of multivariate functions, that learns from a small set of perspective views a function mapping any viewpoint to a standard view, and a network equivalent to this scheme will 'recognize' the object on which it was trained from any viewpoint.

Three-dimensional object recognition is viewpoint dependent

The human visual system is faced with the computationally difficult problem of achieving object constancy: identifying three-dimensional (3D) objects via two-dimensional (2D) retinal images that may

View-dependent object recognition by monkeys